The Dynamics of State Policy Liberalism, 1936–2014 Devin Caughey Christopher Warshaw

The Dynamics of State Policy Liberalism, 1936–2014
Devin Caughey∗
Department of Political Science
MIT
Christopher Warshaw†
Department of Political Science
MIT
First draft: March 5, 2014
This draft: March 4, 2015
Abstract
Applying a dynamic latent-variable model to data on 148 policies collected over eight
decades (1936–2012), we produce the first yearly measure of the policy liberalism of U.S.
states. Our dynamic measure of state policy liberalism marks an important advance
over existing measures, almost all of which are purely cross-sectional and thus cannot be
used to study policy change. We find that, in the aggregate, the policy liberalism of U.S.
states steadily increased between the 1930s and 1970s and then largely plateaued. The
policy liberalism of most states has remained stable in relative terms, though several
states have shifted considerably over time. We also find surprisingly little evidence
of multidimensionality in state policy outputs. Our new estimates of state policy
liberalism have broad application to the study of political development, representation,
accountability, and other important issues in political science.
We appreciate the excellent research assistance of Melissa Meek, Kelly Alexander, Aneesh Anand, Ti↵any
Chung, Emma Frank, Jose↵ Kolman, Mathew Peterson, Steve Powell, Charlotte Swasey, Lauren Ullmann,
and Amy Wickett. We also appreciate the willingness of Frederick Boehmke and Carl Klarner to generously
share their data. We are grateful for research support from the Dean of the School of Humanities, Arts, and
Social Sciences at MIT. All mistakes, however, are our own.
∗
Assistant Professor, Department of Political Science, Massachusetts Institute of Technology,
[email protected]
†
Assistant Professor, Department of Political Science, Massachusetts Institute of Technology, [email protected]
“Change,” Chandler et al. (1974, 108) noted four decades ago, “is both methodologically
and substantively critical for any theory of policy.” This is true of both of the determinants of
government policies, such as shifts in public mood or changes in the eligible electorate (e.g.,
Stimson, MacKuen, and Erikson 1995; Husted and Kenny 1997), and of policy feedback on
political and social outcomes (e.g., Wlezien 1995; Campbell 2012). Theories of all these
phenomena rely explicitly or implicitly on models of policy change. Moreover, many of
the most ambitious theories focus not on individual policies or policy domains, but on the
character of government policy as a whole. In short, most theories of policymaking are
both dynamic and holistic: they are concerned with changes in the general orientation of
government policy.
Unfortunately, the literature on U.S. state politics, perhaps the most vibrant field for
testing theories of policymaking, relies almost exclusively on policy indicators that are either
measured at a single point in time (e.g., Wright, Erikson, and McIver 1987) or else cover
only a partial subset of state policy outputs (e.g., Besley and Case 2003).1 Static measures
are poorly suited to studying causes of policy change over time (Lowery, Gray, and Hager
1989; Ringquist and Garand 1999; Jacoby and Schneider 2009). And while domain-specific
measures may provide useful summaries of some aspects of state policy, such as welfare
spending (Moffitt 2002) or gay rights (Lax and Phillips 2009a), they are at best imperfect
proxies for what is often the outcome of interest, the overall orientation of state policy.
In this paper, we develop a holistic yearly summary of the ideological orientation of state
policies, which we refer to as state policy liberalism. This measure is based on a unique
dataset of 148 policies, which covers nearly eight decades (1936–2014) and includes policy
domains ranging from labor regulation and civil rights to gun control and gay rights.2 Based
on these data, we estimate policy liberalism in each year using a dynamic Bayesian latent1. To our knowledge, the only existing holistic yearly summary of state policies is Jacoby and Schneider’s
(2009) measure of particularistic versus collective state spending priorities between 1982 and 2005. As we
discuss below, our measures di↵er substantially in time coverage, conceptual interpretation, and the data
used to construct them.
2. Both the policy data and our policy liberalism estimates will be made available to the public upon
publication of this article.
1
variable model designed for a mix of continuous, ordinal, and dichotomous policy indicators.
This measurement model enables us to make use of many indicators of policy liberalism,
thus substantially reducing measurement error on the estimates of our construct of interest.
Despite the disparate policy domains covered by our dataset, allowing for additional
latent policy dimensions does little to improve the predictive accuracy of the model. This
suggests that contrary to previous claims (e.g., Sorens, Muedini, and Ruger 2008), a single
latent dimension suffices to capture the systematic variation in state policies. Consistent
with this conclusion, our dynamic measure is highly correlated with existing cross-sectional
measures of state policy liberalism as well as with issue-specific ideological scales.
Substantively, we find that while U.S. states as a whole have drifted to the left (that
is, they have increasingly adopted liberal policies), most have remained ideologically stable in relative terms. Across our entire time series, the most conservative states are in the
South, whereas California, New York, Massachusetts, and New Jersey are always among the
most liberal. The relative policy liberalism of a few states, however, has changed substantially. Several Midwestern and Mountain states have become considerably more conservative
relative to the rest of the nation, whereas most of the Northeast has become more liberal.
Our new dynamic estimates can be used to study a wide variety of possible questions,
many of which are not easily investigated using cross-sectional measures. Potential topics
of study include the short- and long-term determinants of policy outputs, such as economic
development, political institutions, mass policy preferences, and electoral outcomes. Policy
liberalism could also be used as an independent variable, as a means of examining policy
feedback or other consequences of policy change. These measures thus o↵er new research avenues onto political development, representation, accountability, and other important issues
in political science.
The remainder of the paper is organized as follows. We begin by defining the concept of
policy liberalism and situating it in the literature on U.S. state politics and policy. Next, we
describe our policy dataset, our measurement model, and our yearly estimates of state policy
2
liberalism. We then provide evidence for the validity of our measure. We show that it is
highly correlated with existing measures of policy liberalism and related concepts, and that a
one-dimensional scale adequately accounts for systematic policy variation across states. The
penultimate section discusses potential applications of our measure, illustrating its usefulness
with an analysis of the policy e↵ects of voter registration laws. The final section concludes.
Measuring State Policies
Studies of state policy generally employ one of two measurement strategies: they either consider policy separately using policy-specific indicators, or they construct composite measures
intended to summarize the general orientation of state policies within or across domains
(Jacoby and Schneider 2014, 568). Among studies in the first camp, some have focused on
whether or not states have particular policies. Lax and Phillips (2009a), for example, examine the representational congruence between a series of dichotomous state gay-rights policies
and state opinion majorities. Other studies have employed continuous policy-specific indicators, such as welfare expenditures (Husted and Kenny 1997), tax rates (Besley and Case
2003), or minimum wages (Leigh 2008), which potentially have greater sensitivity to differences between states. Whether dichotomous or continuous, policy-specific measures are
appropriate when the research question is limited to a particular policy area. But they are
suboptimal as summary measures of the general orientation of state policies, though this is
how they are often used.3
For this reason, a number of scholars have sought to combine information from multiple
policies, using factor analysis or other dimension-reduction methods to summarize them in
terms of one or more dimensions of variation. Dimension reduction has several advantages
over policy-specific measures. First, from a statistical point of view, using multiple indica3. Lax and Phillips (2009a, 369) claim that “using. . . policy-specific estimates” allows them to “avoid
problems of inference that arise when policy and opinion lack a common metric.” On a policy-by-policy basis
this is probably true. But evaluating congruence on state policy in general, or even just in the domain of
gay rights, requires that the policy-specific estimates of congruence be weighted or otherwise mapped onto a
single dimension. Thus, dimension reduction must occur at some point, whether at the measurement stage
or later in the analysis.
3
tors for a latent trait usually reduces measurement error on the construct of interest, often
substantially (Ho↵erbert 1966; Ansolabehere, Rodden, and Snyder 2008). Secondly, many
concepts require multiple indicators to adequately represent the full content or empirical domain of the concept. For example, the concept of liberalism, in its contemporary American
meaning, encompasses policy domains ranging from social welfare to environmental protection to civil rights. A measure of liberalism based on only a subset of these domains would
thus fare poorly in terms of content validation (Adcock and Collier 2001, 538–40). A final
benefit is parsimony. If a single measure can predict variation in disparate domains, then we
have achieved an important desideratum of social science: “explaining as much as possible
with as little as possible” (King, Keohane, and Verba 1994, 29).
Di↵erent works have identified di↵erent traits or dimensions underlying state policies.
Walker (1969), for example, creates an “innovation score” that captures the speed with
which states adopt new programs. Sharkansky and Ho↵erbert (1969) identify two latent
factors that structure variation in state policies, as do Sorens, Muedini, and Ruger (2008).
Hopkins and Weber (1976) uncover a total of five. But primarily the state politics literature
has focused on a single left–right policy dimension (e.g., Ho↵erbert 1966; Klingman and
Lammers 1984; Wright, Erikson, and McIver 1987; Gray et al. 2004). As a number of studies
have confirmed, states with minimal restrictions on abortion tend to ban the death penalty,
regulate guns more tightly, o↵er generous welfare benefits, and have progressive tax systems,
and vice versa for states with more restrictive abortion laws. Following Wright, Erikson, and
McIver (1987), we label this dimension policy liberalism.
What is policy liberalism? We conceptualize liberalism not as a logically coherent ideology, but as a set of ideas and issue positions that, in the context of American politics, “go
together” (Converse 1964). Relative to conservatism, liberalism involves greater government
regulation and welfare provision to promote equality and protect collective goods, and less
government e↵ort to uphold traditional morality and social order at the expense of personal
autonomy. Conversely, conservatism places greater emphasis on the values of economic free-
4
dom and cultural traditionalism (e.g., Ellis and Stimson 2012, 3–6). Although the definitions
of liberalism and conservatism have evolved over time, with civil rights and then social issues
becoming more salient relative to economics (Ladd 1976, 589–93), these ideological cleavages
have existed in identifiable form since at least the mid-20th century (Schickler 2013; Noel
2014).
There are several things to note about this definition of policy liberalism. First, it is
comprehensive, in that it covers most if not all domains of salient policy conflict in American
domestic politics.4 This is not to say that policy liberalism explains all variation in state
policy, or that all policies are equally structured by this latent dimension. But it is a concept
that attempts to summarize, holistically, all the policy outputs of a state. Second, we define
policy liberalism solely in terms of state policies themselves. By contrast, some previous
measures (e.g., Sharkansky and Ho↵erbert 1969; Hopkins and Weber 1976) incorporate societal outcomes like infant mortality rates and high school graduation rates, muddying the
distinction between government policies and socio-economic conditions (Sorens, Muedini,
and Ruger 2008).
A final characteristic of our conceptualization of policy liberalism, which is particularly
crucial for our purposes, is that it is dynamic. Unlike, say, state political culture (Elazar
1966), which changes slowly if at all, policy liberalism can and does vary across time in
response to changes in public opinion, partisan control, and social conditions. Defining policy
liberalism as a time-varying concept is hardly controversial, but it does conflict with previous
operationalizations of this concept, all of which are cross-sectional. Cross-sectional measures
are problematic for two reasons. First, many are based on data from a long time span—over
a decade, in the case of Wright, Erikson, and McIver (1987)—averaging over possibly large
year-to-year changes in state policy (Jacoby and Schneider 2001). More importantly, crosssectional measures preclude the analysis of policy change, which not only is theoretically
limiting, but also inimical to strong causal inference since the temporal order of the variables
4. We do not include foreign policy in the domain of policy liberalism because states typically do not make
foreign policy.
5
cannot be established (Lowery, Gray, and Hager 1989; Ringquist and Garand 1999).
To our knowledge, the only existing time-varying measure that provides a holistic summary of state policy outputs is the measure of policy spending priorities developed by Jacoby
and Schneider (2009).5 This measure, available annually between 1982 and 2005, is estimated
with a spatial proximity model using data on the proportions of state budgets allocated to
each of nine broad policy domains (corrections, education, welfare, etc.). Jacoby and Schneider interpret their measure as capturing the relative priority that states place on collective
goods versus particularized benefits, an important concept in the theoretical literature on
political economy (e.g., Persson and Tabellini 2006) as well as in empirical work on state
politics (e.g., Gamm and Kousser 2010).
Despite both being holistic yearly policy measures, policy liberalism and policy priorities
di↵er in important ways. As Jacoby and Schneider emphasize, policy liberalism and policy
priorities are conceptually distinct; indices of policy liberalism “simply do not measure the
same thing” as their policy priorities scale (2009, 19). For example, the policy priorities scale
is not intended to capture “how much states spend” but rather “how states divide up their
yearly pools of available resources” (Jacoby and Schneider 2009, 4). Consequently, variation
in the size of government, which lies at the heart of most liberal–conservative conflict (e.g.,
Meltzer and Richard 1981; Stimson 1991), is orthogonal to their measure. Another salient
di↵erence is that the policy priorities scale is based solely on state spending data. This
endows their measure with a direct and intuitive interpretation, but at the cost of excluding
taxes, mandates, prohibitions, and other non-spending policies that shape the lives of citizens
in equally important ways. Our policy liberalism measure resolves this trade-o↵ di↵erently,
emphasizing broad policy coverage at the possible expense of intuitive interpretation.
In summary, there is no existing time-varying measure of state policy liberalism, one of the
central concepts of state politics. Nearly all existing summaries of state policy orientations
are cross-sectional. Those that are dynamic either examine policy liberalism in a particular
5. For a cross-sectional implementation of this measure, see Jacoby and Schneider (2001).
6
policy area or, in the case of Jacoby and Schneider’s policy priorities scale, measure a di↵erent
concept entirely. Thus what is required is a measurement strategy that summarizes the global
ideological orientation of state policies using time-varying data that capture the full empirical
domain of policy liberalism.
Policy Data
As Jacoby and Schneider (2014) observe, composite measures of policy liberalism risk tautology if they are derived from policy indicators selected for their ideological character.
Although the resulting scale may be a valid measure of policy liberalism, selection bias in
the component indicators undermines any claim that state policies vary along a single dimension. For this reason, we sought to make our dataset of state policies as comprehensive
as possible, so as to allow ideological structure to emerge from the data rather than imposing
it a priori. Given resource constraints and data limitations, we cannot claim to have constructed a random sample of the universe of state policies (if such a thing is even possible).
We are confident, however, that our dataset of 148 distinct policies is broadly representative
of the policy outputs of states across a wide range of domains. (For complete details on the
policies in our dataset, see the online appendix accompanying this article.)
To be included in our dataset, a policy had to meet the following criteria. First, it had
to be a policy output rather than a policy outcome (i.e., an aspect of the social environment
a↵ected by policy) or a government institution (i.e., one of the basic structures or rules of
the government). For example, we excluded state incarceration and infant-mortality rates,
which we considered outcomes. We also excluded indicators for whether states had particular
legislative rules or government agencies, which we classified as institutions.6 Second, the
policy had to be politically salient. To identify salient policies, we canvassed books and
articles on state politics, legal surveys of state policies, state party platforms, governors’
biographies, state-specific political histories, and government and interest-group websites.
6. The dataset used in this paper excludes electoral policies as well. We do this for the pragmatic reason
that scholars may want to use our measure to examine the e↵ect of such policies.
7
Third, the policies had to be comparable across all states. Many environmental, parks, and
farm policies, for example, are not comparable across states due to fundamental di↵erences
in state geography (e.g., coastal versus non-coastal). Some policies we normalized by an
appropriate baseline to make them more comparable.7 Finally, in keeping with our focus
on dynamics, data on a given policy had to be available in comparable form in at least five
di↵erent years.
The actual policy data themselves were obtained from many di↵erent sources, including
government documents, the Book of the States, interest-group publications, and various
secondary sources.8 Over four-fifths of the policies are ordinal (primarily dichotomous), but
the 26 continuous variables provide disproportionate information because they di↵erentiate
more finely between states.9 The policy domains covered by the dataset include
• abortion (e.g., parental notification requirements for minors)
• criminal justice (e.g., the death penalty)
• drugs and alcohol (e.g., marijuana decriminalization)
• education (e.g., per-pupil education spending; ban on corporal punishment)
• the environment (e.g., protections for endangered species)
• civil rights (e.g., fair employment laws; gay marriage)
• gun control (e.g., handgun registration)
• labor (e.g., right-to-work laws)
• social welfare (e.g., AFDC/TANF benefits)
• taxation (e.g., income tax rates)
and miscellaneous other regulations, such as fireworks bans and bicycle helmet laws.
To validate the comprehensiveness of our dataset, we can compare its coverage to other
datasets that were constructed for di↵erent purposes. For example, our policies cover 17
7. We adjusted all monetary expenditure and welfare benefit policies into 2012 dollars. We also adjusted
for cost-of-living di↵erences between states (Berry, Fording, and Hanson 2000).
8. In general, we tried to obtain primary sources for each policy indicator. When this proved impossible,
we obtained multiple secondary sources to corroborate the information about each policy in our database.
9. We standardized each continuous policy to ensure that the scales were comparable across policy areas.
8
of the 20 non-electoral policy areas contained in Sorens, Muedini, and Ruger’s (2008) state
policy database. Similarly, seven of the eight policy categories in the National Survey of State
Laws, a lengthy legal compendium of “the most-asked about and controversial” state statutes,
are represented in our dataset (Leiter 2008, xii).10 Our data also include 40 of the 56 policy
outputs in Walker’s (1969) policy innovation dataset and 21 of the 34 non-electoral policies
examined by Lax and Phillips (2011).11 The overlap between these last three datasets and
ours is particularly significant, because none of the three were constructed for the purpose of
studying the ideological structure of state policies. Even Sorens, Muedini, and Ruger (2008),
who do analyze policy in ideological terms, conceive of state policies as varying along two
dimensions. In sum, our dataset, while not a random sample of the universe of policies, is
broadly representative of available data on the salient policy activities of U.S. states.
Measurement Model
We use the policy dataset described above to construct yearly measures of state policy
liberalism. Like most previous work on the subject, we treat policy liberalism as a latent
variable whose values can be inferred from observed policy indicators. Our latent-variable
model (LVM), however, o↵ers several improvements over previous measurement strategies,
most of which have relied on factor analysis applied to cross-sectional data. First, we use
a Bayesian LVM, which unlike classical factor analysis provides straightforward means of
characterizing the uncertainty of the latent scores and also easily handles missing data by
imputing estimates on the fly (Jackman 2009, 237–8). Second, most of our policy indicators
are dichotomous variables, a poor fit for a factor-analytic model, which assumes that the
observed indicators are continuous. We therefore follow Quinn (2004) and specify a mixed
LVM that models continuous indicators with a factor-analytic model and ordinal (including
10. The categories are Business and Consumer, Criminal, Education, Employment, Family, General Civil,
Real Estate, and Tax. There are no real estate laws in our dataset because we could not locate comparable
time-varying data on these laws.
11. The remaining policies are missing either because time-varying data were not available or because the
policies are not sufficiently comparable across states.
9
dichotomous) variables with an item-response model. Third, our measurement model is
dynamic, both in that it allows policy liberalism to vary by year and in that it specifies a
dynamic linear model that links the measurement model between periods.
We parameterize policy liberalism as a latent trait ✓st that varies across states and years.
For each state s and year t, we observe a mix of J continuous and ordinal policies, denoted
yst = (y1st , . . . , yjst , . . . , yJst ), whose distribution is governed by a corresponding vector of
⇤
⇤
latent variables yst
. We model yst
as a function of policy liberalism (✓st ) and item-specific
parameters ↵t = (↵1t , . . . , ↵jt , . . . , ↵Jt ) and
= ( 1, . . . ,
⇤
yst
⇠ NJ ( ✓st
↵t ,
j, . . . ,
J ),
),
(1)
is a J ⇥ J
where NJ indicates a J-dimensional multivariate normal distribution and
covariance matrix. In this application, we assume
to be diagonal, but this assumption
could be relaxed to allow for correlated measurement error across variables. Note that ↵jt ,
which is analogous to the “difficulty” parameter in the language of item-response theory,
varies by year t, whereas the “discrimination”
j
is assumed to be constant across time.
We accommodate data of mixed type via the function linking latent and observed vari⇤
⇤
ables. If policy j is continuous, we assume yjst
is directly observed (i.e., yjst = yjst
), just
as in the conventional factor analysis model. If policy j is ordinal, we treat the observed
⇤
yjst as a coarsened realization of yjst
whose distribution across Kj > 1 ordered categories is
determined by a set of Kj + 1 thresholds ⌧j = (⌧j0 , . . . , ⌧jk , . . . , ⌧j,Kj ). Following convention,
we define ⌧j0 ⌘
1, ⌧j1 ⌘ 0, and ⌧jKj ⌘ 1, and we set the diagonal elements of
that cor-
respond to ordinal variables equal to 1. As in a ordered probit model, yjst falls into category
k if and only if ⌧j,k
⇤
that yjst
⇠ N( j ✓st
Pr(⌧j,k
1
1
⇤
< yjst
 ⌧jk . Thus for ordinal variable j, the conditional probability
↵jt , 1) is observed as yjst = k is
⇤
< yjst
 ⌧jk |
j ✓st
⇤
↵jt ) = Pr(yjst
 ⌧jk |
= (⌧jk
[ j ✓st
10
j ✓st
↵jt ])
↵jt )
(⌧j,k
⇤
Pr(yjst
 ⌧j,k
1
[ j ✓st
1
|
↵jt ]),
j ✓st
↵jt )
(2)
where
is the standard normal CDF (Fahrmeir and Raach 2007, 329). In the dichotomous
case, where there are Kj = 2 categories (“0” and “1”), the conditional probability that yjst
falls in the second category (i.e., “1”) is
⇤
Pr(⌧j1 < yjst
 ⌧j2 |
j ✓st
↵jt ] =
=
(⌧j2
[ j ✓st
( j ✓st
↵jt ])
(⌧j1
[ j ✓st
↵jt ])
↵jt ),
(3)
which is identical to the conventional probit item-response model (Quinn 2004, 341).
We allow the ↵jt to vary by year to account for the fact that many policies (e.g., segregation laws) trend over time towards universal adoption or non-adoption. The simplest
way to deal with this problem is to estimate the difficulty parameters anew in each year. A
more general approach, however, which pools information about ↵jt over time, is to model
the evolution of the ↵jt with a dynamic linear model, or DLM (West and Harrison 1997;
Jackman 2009, 471–2). In this application we use a local-level DLM, which models ↵jt using
a “random walk” prior centered on ↵j,t 1 :
↵jt ⇠ N(↵j,t 1 ,
2
↵ ).
(4)
If there is no new data for an item in period t, then the transition model in Equation 4 acts
as a predictive model, imputing a value for ↵jt (Jackman 2009, 474). The transition variance
2
↵
controls the degree of smoothing over time. Setting
↵jt separately each year, and
2
↵
2
↵
= 1 is equivalent to estimating
= 0 is the same as assuming no change over time. We take
the more agnostic approach of estimating
2
↵
from the data, while also allowing it to di↵er
between continuous and ordinal variables.
The parameters in an LVM cannot be identified without restrictions on the parameter
space (e.g., Clinton, Jackman, and Rivers 2004). In the case of a one-dimensional model,
the direction, location, and scale of the latent dimension must be fixed a priori. We identify
the location and scale of the model by post-processing the latent measure of state policy
11
liberalism to be standard normal. For the prior on the innovation parameter
↵,
we use a
half-Cauchy distribution with a mean of 0 and a scale of 2.5 (Gelman 2006). The difficulty
and discrimination parameters are drawn from normal distributions with a mean of 0 and
a standard deviation of 10. We fix the direction of the model by constraining the sign of
a small number of the item parameters (Bafumi et al. 2005).12 We further constrain the
polarity by assigning an informed prior to the policy measure for four states in year t = 0
(Martin and Quinn 2002).13 We estimated the model using the program Stan, as called
from R (Stan Development Team 2013; R Core Team 2013).14 Running the model for 1,000
iterations (the first 500 used for adaptation) in each of 4 parallel chains proved sufficient to
obtain satisfactory samples from the posterior distribution.
Estimates of State Policy Liberalism
Estimating our measurement model using the policy data described earlier produces a measure of the policy liberalism of each state in each year 1936–2014. When interpreting these
estimates, one should bear in mind that the model allows the difficulty parameters ↵t to
evolve over time. As a result, aggregate ideological shifts common to all states will be partially assigned to the policy difficulties. Since states did adopt increasingly liberal policies
over this period, the model partially attributes this trend to the increasing difficulty of conservative policies (and increasing “easiness” of liberal ones). If we modify the model so as to
hold the item difficulties constant over time, the policies of all U.S. states are estimated to
12. Specifically, we constrain continuous measures of state spending to have a positive discrimination parameter, which implies that more liberal states spend more money. We also constrain the polarity of four
dichotomous items. The discrimination of ERA ratification and prevailing wage laws are constrained to be
positive, while the discrimination of right to work laws and bans on interracial marriage are constrained to
be negative.
13. Note that we started the model in 1935 (t = 0) and discarded the first year of estimates. As a result, the
informed priors on ✓ for four states in year t = 0 have little e↵ect on the estimates of state policy liberalism
that we report in our analysis. We assign a N(1, 0.22 ) prior on ✓s0 to New York and Massachusetts, and a
N( 1, 0.22 ) prior for Georgia and South Carolina. Other states are given di↵use priors for ✓st .
14. Stan is a C++ library that implements the No-U-Turn sampler (Ho↵man and Gelman, Forthcoming), a
variant of Hamiltonian Monte Carlo that estimates complicated hierarchical Bayesian models more efficiently
than alternatives such as BUGS.
12
1940
1975
2010
Figure 1: The geographic distribution of government policy liberalism in 1940, 1975, and
2010. Darker shading indicates liberalism; lighter shading indicates conservatism. The estimates have been centered and standardized in each year to accentuate the shading contrasts.
have become substantially more liberal, especially between the 1930s and 1970s.15 We use a
time-varying model instead because it helps avoid the interpretational difficulties of assuming that policies have the same substantive meaning across long stretches of time. The price
of this flexibility is that states’ policy liberalism scores are comparable over time primarily
in a relative sense.
Figure 1 maps state policy liberalism in 1940, 1975, and 2010. As is clear from this figure,
the geographic distribution of policy liberalism has remained remarkably stable, despite huge
changes in the distribution of mass partisanship, congressional ideology, and other political
variables over past seven decades. Throughout the period, Southern states had the most
conservative policies. This holds not only on civil rights, but on taxes, welfare, and a host
of social issues. By contrast, the most liberal states have consistently been in the Northeast,
Pacific, and Great Lakes regions. New York, for example, has consistently had the most
liberal tax and welfare policies in the nation, and it was also among the first states to adopt
liberal policies on cultural issues such as abortion, gun control, and gay rights.
The overall picture of aggregate stability, however, masks considerable year-to-year fluctuation in policy liberalism as well as major long-term trends in certain states. These details
can be discerned more easily in Figure 2, which plots the yearly time series of individual states
15. In these years, U.S. states expanded their welfare responsibilities and tax bases while loosening a variety
of social restrictions. This aggregate trend towards more liberal policies largely ceased after 1980.
13
3
CA NJ
HI
CT NY
MA
State Goverment Policy Liberalism
2
MD RI
VT
ME
OR
WA DE
NM MN
1
0
−1
CA NJ NY
MA
RI CO
WI
MI WAOH
PA MN
IL CT
UT
ME
OR MT
IN IA NH
ID AZ
NM NE KS
NV MDND
SD
MO VT
LA
KY
WY
IL
WI
PA
IA
MI MT
CO
NV
AK
OH
NEWV
IN
KY
TX
MO
VA
KS AZ
TN LA FL
WY
SD UT OK
ID
OK
DE
NH
TX NC
AL VA
TN FL WV
SC
GA
ND
NC
AR
−2
AL
MS
AR
SC GA
MS
1936
1950
1975
2000
2014
Year
South
Mountain West
Midwest
Northeast
Pacific Coast
Figure 2: State government policy liberalism, 1936–2014. The thicker black line tracks the
mean in each year, and the colored lines indicate the means in five geographic regions.
14
between 1936 and 2014. Due to explicit policy revisions as well as to policy “drift” relative to
other states, policy liberalism can change substantially between years, though cross-sectional
di↵erences between states are generally much larger than within-state changes. The variance
across states has also increased over time, possibly due to growing geographic polarization.
Figure 2 also shows that not all states have been ideologically stable. The policies of
Northeastern states became steadily more liberal over this time period. Whereas states like
Delaware, Maryland, and Vermont were once more conservative than average, by 2014 all
three had joined most of the rest of the Northeast in the top quartile of liberalism. Their
early adoption of gay marriage and other rights for homosexuals, for example, contrasts with
their slowness in passing racial anti-discrimination laws in the 1950s and 1960s. The welfare
benefits and regulatory policies of these states exhibited a similar liberalizing trajectory.
Several Midwestern, Mountain, and Southern states have followed the opposite trajectory.
Idaho, for example, became much more conservative over this period. In the 1930s–1950s,
Idaho actually had some of the most generous welfare benefits in the nation, but by the
early 2000s they were among the least generous. Louisiana too has shifted substantially to
the right. In the 1930s, Louisiana’s welfare benefits were the most generous in the South
and roughly equivalent to those of several Northern states, but they gradually become less
generous over the next few decades. Louisiana also waited longer than any other Southern
state to pass a durable right-to-work law, but it finally did so in 1976.16
These states’ shifts in policy liberalism track the evolution of their presidential partisanship. For instance, in the presidential election of 1936, the first year in our dataset, Maine,
Vermont, and New Hampshire were the three most Republican states in the nation, but by
2012 all three (especially Vermont) were more Democratic than average. The opposite is true
of the Mountain West, which transformed from Democratic-leaning to solidly Republican.
On the whole, the 2010 map in Figure 1 matches contemporaneous state partisanship much
16. Louisiana passed a right-to-work law in 1954 but repealed it in 1956, when the populist Long faction of
the Democratic Party recaptured control of state government (Canak and Miller 1990). The unusual power
of this faction, forged by Governor and Senator Huey Long in the late 1920s, may help explain Louisiana’s
anomalously (for the region) liberal state policies in that era (Key 1949, 156–82).
15
better than the earlier maps, primarily because the South’s shift to the Republicans finally
aligned its partisanship to match its consistently conservative state policies.
Measurement Validity
Having illustrated the face validity of the policy liberalism estimates, we now conduct a
more systematic validation of our measure. We begin with convergent validation (Adcock
and Collier 2001), documenting the very strong cross-sectional relationships between our
estimates’ and existing measures of policy liberalism. We then turn to construct validation,
demonstrating that our policy liberalism scale is also highly correlated with measures of
theoretically related concepts, such as presidential partisanship. Finally, we show that our
policy liberalism scale is strongly related to domain-specific policy measures, and that the
predictive fit of the model barely increases if a second dimension is added to the measurement model. Overall, this evidence corroborates our claim that a one-dimensional model
adequately captures the systematic variation in state policies, and that this dimension is
properly interpreted as policy liberalism.
Convergent Validation
If our estimates provide a valid measure of policy liberalism, they should be strongly related
to other (valid) measures of the same concept. Since ours is the first time-varying measure
of state policy liberalism, we must content ourselves with examining the cross-sectional
relationship between our measure and ones developed by other scholars at various points in
time. Figure 3 plots the cross-sectional relationships between our measure of policy liberalism
and six existing measures:
• “liberalness”/“welfare orientation” rank circa 1957 (Ho↵erbert 1966)17
• welfare-education liberalism in 1962 (Sharkansky and Ho↵erbert 1969)18
17. This index is based on mean per-recipient expenditures for 1952–61 for aid to the blind, old age
assistance, unemployment compensation, expenditure for elementary and secondary education, and aid to
dependent children. We compare Ho↵erbert’s (1966) scale with our measure of state policy liberalism in
1957 since this is the midpoint of the years he includes in his index.
18. This index is based on about twenty education and welfare policies. Note, however, that this index
16
MS
40
30
20
10
SC AL AR
NC
TN
VA WV
KY
GA
FL
MO ME
VT
SDTX
LA
NM
MD
DE NE IN
OK
ID OH
UT
MT
AZ ND NH
IA
KS
NV
WY
r=−0.76
WA
PA
RI
MI
ILCO
MN OR
WI
CA CT NJ
0
−2
−1
0
MA
NY
1
Policy in 1962 (Shar. & Hoff., 1969)
Policy in ~1957 (Hofferbert, 1966)
1957
50
1962
2
CA
WI MA
MN
IL
NH
NJ
IAND
OK KS
MIWA
OR
CO CT
NE
RI
OH
WY NM
INUT
PA
ID
SD
MD MT
MO
DE AZ
NV
VT LAME
FL TX
KY
VA
NC
AR
WV
GA
SC
TN
AL
r=0.85
1
0
−1
−2
MS
2
−2
−1
1973
1
0
−1
SC
−2
ND
TX
FL
LA
VA
NC
GA
TN
AL
NJ
CA
CT
OR
WI
CO MN PA MI
RI
WA
IL
NH VT MD
IA
KS
ID
ME
MTOH
DE
WY
NM
NE
KY
SD
UT WV
IN
MO OK
NV
AZ
−2
NY
2
r=0.9
OR
WI
MI
DEPACT
MD
VT
MN
OH MT
CO IA
IL
WA
KS WV
ME
NH
KY
FL
ND
MO UT
TX
WY
VA
TN ID
NC
OK
NM
GA
LA
AZ SD
IN
SC
AR
1
0
−1
AR
MS
AL
MS
−1
0
1
2
−1
0
Policy Liberalism
OR
MT
NM DE
MD
WV
ME
IL NH WA
−2.5
−5.0
CO
MO
OH
NE
KS IN
TXAZ
KY
OKVA NV
ID
UT
MS AL
TN
AR
WY
GA
NC
ND LA
FL
SD
SC
−2
−1
0
MI
IAPA
RI
NJ
WI
1
Policy in 2006 (Sorens, et al, 2008)
Policy in 2000 (Gray, et al, 2004)
0.0
NY
VT MA
MN CT
r=0.88
2.5
NJ
RI
1
2
2006
CA
5.0
MA
CA
Policy Liberalism
2000
7.5
1
1980
NY
MA
r=0.9
0
Policy Liberalism
Policy in ~1980 (EWM, 1993)
Policy in ~1973 (Klingman & Lammers, 1984)
Policy Liberalism
2
NY
MA NYCA
NJ
10
r=0.84
RI
MD
HI
IL
5
MI
OH
PA
OR
IA
FL CO
NH
WV AK WI
NV
KY
VA
AZ
LA
MO
MTNM
OK
TX KSIN NE
TN
SD
ID
ND UT
WY
NC
0
−5
CT
DE
MS
2
SC
AL GA
AR
−2
Policy Liberalism
−1
0
1
WA
MN MEVT
2
Policy Liberalism
Figure 3: Convergent validation: relationships between our policy liberalism estimates and
six existing measures. Fitted lines indicate loess curves.
• policy liberalism circa 1973 (Klingman and Lammers 1984)19
• policy liberalism circa 1980 (Wright, Erikson, and McIver 1987)20
• policy liberalism in 2000 (Gray et al. 2004)21
also includes several social outcomes, such as school graduation rates.
19. This index is based on data measured at a variety of points between 1961 and 1980 on state innovativeness, anti-discrimination policies, monthly payments for Aid to Families with Dependent Children (AFDC),
the number of years since ratification of the Equal Rights Amendment for Women, the number of consumeroriented provisions, and the percentage of federal allotment to the state for Title XX social services programs
actually spent by the state. We compare Klingman and Lammers’s (1984) scale with our measure of state
policy liberalism in 1973 since this is the midpoint of the years they include in their index.
20. This measure is based on state education spending, the scope of state Medicaid programs, consumer
protection laws, criminal justice provisions, whether states allowed legalized gambling, the number of years
since ratification of the Equal Rights Amendment for Women, and the progressivity of state tax systems.
We compare Wright, Erikson, and McIver’s (1987) scale with our measure of state policy liberalism in 1980
since this is roughly the midpoint of the years they include in their index.
21. This index is based on state firearms laws, state abortion laws, welfare stringency, state right-to-work
laws, and the progressively of state tax systems.
17
• policy liberalism in 2006 (Sorens, Muedini, and Ruger 2008)22
Each panel plots the relationship between our policy liberalism estimates (horizontal axis)
and one of the six existing measures listed above. A loess curve summarizes each relationship,
and the bivariate correlation is given on the left side of each panel.
Notwithstanding measurement error and di↵erences in data sources, our estimates are
highly predictive of other measures of policy liberalism. The weakest correlation, 0.76 for
Ho↵erbert (1966), is primarily the result of a few puzzling outliers (Washington, for example,
is the seventh-most conservative state on Ho↵erbert’s measure, whereas Wyoming is the
ninth-most liberal). In addition, all the relationships are highly linear. The only partial
exception is for Sorens, Muedini, and Ruger (2008), whose measure of policy liberalism
does not discriminate as much between Southern states as our measure, resulting in a flat
relationship at the conservative end of our scale.
In short, the very strong empirical relationships between our policy liberalism scale and
existing measures of the same concept provide compelling evidence for the validity of our
measure. It is worth noting that most of the existing scales were constructed explicitly with
the goal of di↵erentiating between liberal and conservative states. Thus their tight relationship with our measure, which is based on a much more comprehensive policy dataset and
was estimated without regard to the ideological content of the policy indicators,23 suggests
in particular that we are on firm ground in calling our latent dimension “policy liberalism.”
Construct Validation
The purpose of construct (a.k.a. “nomological”) validation is to demonstrate that a measure conforms to well-established hypotheses relating the concept being measured to other
concepts (Adcock and Collier 2001, 542–3). One such hypothesis is that the liberalism of
a state’s policies is strongly related to the liberalism of its state legislature, though due to
22. This is the first principal component uncovered by Sorens, Muedini, and Ruger’s (2008) analysis of
over 100 state policies. They label this dimension “policy liberalism” and give the label “policy urbanism” to
the second principal component.
23. This is true except for the hard coding required to identify the latent scale.
18
factors such as legislative gridlock the relationship may not be perfect (e.g., Krehbiel 1998).
To measure legislative liberalism on a common scale, we rely on Shor and McCarty’s (2011)
estimates of the conservatism of members of state legislative lower houses. As Figure 4
demonstrates for presidential years between 1996 and 2008, states with more liberal policies
tend to have more liberal median legislators. Due possibly to the lingering Democratic advantage in Southern state legislatures, the relationship at the conservative end of the policy
spectrum is fairly flat, though by 2008 the relationship had become much more linear. The
correlation between legislative conservatism and policy liberalism has also strengthened over
time, from
0.51 in 1996 to
0.80 in 2008.
Median Legislator in State House (Shor & McCarty, 2011)
1996
0.5
SC
NC
TX
AL
LA
0.0
GA
0.5
IN
KS
KY
OK
FL
TNVA
IA
OH
NH
CA
MI
WV
MD
0.5
ND
MS
AR NC
AL
0.0
UT
0
2004
TX
GA
CT
NY
MA
−1.0
2
−2
−1
0.5
MN
ND
MS
AR
OR
0
SD
IDTX
1
CA
2
AK
OH
WI
AL
TN
NC
WA
NM
MO
AZ
WY VA
KS
FL
LA
KY
0.0
VT
WV
r=−0.69
−0.5
RI
HI
SC
GA
IA NH
PA
IL DE
IN
NV
NM
2008
MI
TN
ME
MD HI
NY
1
WI
MT
NJ
VT
IL
WV
RI
AK
MO
OH
CO
ID AZ
WY VA
OK FL
KS
KY
LA
OR
PANH
DE
IN
MN
WA
NV
MA
SC
TN
WI
IA
MI
r=−0.56
−0.5
NM
−1
OH
VT
MN
−1.0
MT
MO
GA
NC
NJ
MO
AZ
KS
CO
FL KY
TX
LA OKVA
AR
AR
AL
CT
−2
ND
WY
0.0
PAME
DE
AK
SD ID
UT
MS
WI
IL
r=−0.51
−0.5
SC
ID
CO
AZ
MS
2000
AK WA
SD
WY
UT
r=−0.8
−0.5
ME
RI
OR
IL
IN
MT
DE
PA
WV
MI
CO
IANH
ME
WA
MN
MDHI
CT
NY
MA NJ
−1.0
RI
MD HI
CT
−1.0
NY
NJ
CA
CA
−1.5
−1.5
−2
−1
0
1
2
−2
−1
0
1
2
Policy Liberalism
Figure 4: The relationship between state policy liberalism and the conservatism of the median
member of the lower house of the state legislature (Shor and McCarty 2011), 1996–2008.
An analogous pattern of increasing association over time can be seen in an examination
of the relationship between policy liberalism and Democratic presidential vote share. It is
natural to hypothesize that both presidential vote and state policy liberalism are responsive
19
to the party and policy preferences of mass publics and thus should be correlated at the state
level. Since the anomalously Democratic partisanship of the “Solid South” would distort this
relationship, we focus on the non-South only. Even without Southerns states, however, policy liberalism and presidential vote are only weakly related in the early part of the period,
as Figure 5 shows. The correlation jumped to 0.58 in 1960 and continued to increase gradually through 2012, when it reached nearly 0.9. This increasing association between policy
liberalism and presidential vote mirrors the growing alignment of party identification, policy preferences, and presidential vote at the mass level (Fiorina and Abrams 2008, 577–82).
The analysis of presidential vote thus provides further evidence for the validity of our policy liberalism scale. At same time, however, it suggests the limitations of presidential vote
share as a proxy for mass preferences before the 1960s, even in the non-South (contra, e.g.,
Canes-Wrone, Brady, and Cogan 2002).
Finally, we examine the relationship between our policy liberalism measure and its closest analogue, Jacoby and Schneider’s (2009) policy priorities scale. As we emphasize above,
policy liberalism and policy priorities are di↵erent concepts. Moreover, the theoretical relationship between policy liberalism and preference for collective over particularistic spending
is not self-evident. Nevertheless, Jacoby and Schneider convincingly argue that in U.S. states
tend to target particularized policies at needy constituencies. Consistent with that expectation, they find a moderately negative cross-sectional correlation between policy liberalism
and preference for collective goods.
Based on a similar analysis, we too find policy liberalism and policy priorities to be
negatively correlated, on the order of
0.5. As Figure 6 shows, their relationship atten-
uated somewhat between 1982 and 2005. Also, like Jacoby and Schneider (2009, 18–20),
we find that non-linearity in the measures’ relationship contributes to the weak correlation:
their association is much stronger among relatively liberal and particularistic states than on
the conservative/collective-good end of the spectrum. This seems to be driven in part by
Southern states, which always anchor the conservative end of our scale but seem to favor par-
20
1936
70
60
50
1940
NV
AZ
MT
NDORUT WA
WICA
ID MN
NM
MD
CO
WY
MO
OH
NJ
ILPA
MI NY
NEIN CT
RI
IA
MA
DE SD
KS
NH
VT
40
70
60
50
ME
30
40
1952
Democratic Presidential Vote % (Non−South)
40
30
50
MO
DE
RI
PA
WA MA
MN IL
MI
MDNM
OH
CACTNJ NY
AZMT
IN UT
ORWI
NVWY NH CO
IA
ID
ME
SDNE KS
ND
VT
40
30
r=0.44
RI
60
MN
WA
OR
DE
MI
PA
MT CA
NV
SD
IAIN NM
IL RIMA
MD
WY
CO
AZ
OH
ID
NY
ND
WI
CT
NJ
NE UT
KS
NH
VT ME
NY
50
30
r=0.61
60
r=0.72
20
50
40
30
50
40
30
50
RI
SDMN OR
WI
CA
MI
IA CT
ILWA PA NY
DE
MT
OH
ME MD
MO
NM
AK HI NJ
VT
NDNV
CO
NH
IN
AZ
WYNE
KS
UT ID
40
60
50
40
r=0.81
20
−1
0
1
40
30
2
60
50
40
30
r=0.82
1
60
50
40
30
r=0.66
r=0.76
20
2
2012
HI
VT
RI
MA NY
MD
IL
DE
CT CA
WA
MI WI
ORME
NJ
NM
NV
PA
NH MN
IA
CO
OH
IN
MO
MT
NDSD AZ
KS NE
AK
ID
UT
WY
r=0.82
20
−1
0
1
r=0.71
MA RI
NY
HI
ME VT
IL
MNMD CTNJ
DE
MI WI
WA
CA
IAOR
NH PA
NM
MO OH
AZNV
MT
SD CO
IN
ND
WY
KS
AK
UT ID NE
2008
MA
VTRI NY
MD CT
ILDE ME
HI NJ CA
WA
ORMN
MI WI
PA
IANHNM
0
UT
1996
70
20
IN MT
SD
KS
AK
ND
NE
ID
WY
UT
−1
40
20
MA
VTRI
NY CA
IL WA
MD
MN
HI
MONMDE
MEMIOR
PA
IA CT
CO
WI
NV OH
MT
NJ
NH
AZ
SD
KS IN
WY
AK
ND
ID
NE
UT
70
20
RI
MN HI
MD
MA
DE
NY
ME
WI
MI
MO
PA
ILVTOR
CT
OHIA
WA
NJ
CA
INNM
KS CO
MT
SD NH
AK
AZ
WY
ND NV
IDNE
50
30
r=0.54
30
r=0.67
NVOH
CO
MO
AZ
60
1992
70
50
r=0.58
1980
70
60
40
20
MN RI MA
MD
DE
NY
MO
PA
HI
WI
OH
OR
ME
IA
SD NM IL
CA NJ
WA
MI
ND NVKS
IN MTCT
NH
COVT
AZ
WY
NE ID AK
UT
20
RI
IA MN HI
MA
OR
NY
WI
WA
IL
PA
MD
VT
CA
MO
NMMT CT
SD CO
MI
DEME
ND KS OH
NJ
NE
AZ
NV
WYIN
ID NHAK
UT
NY
70
2004
RI
MA
NY
MDHICTNJ
ILDE
CA
VT
WA
MI ME
PA
WI ORMN
NMIANH
MO
NVOH
AZ CO
IN
SDKS
ND
NE MT
AK
ID
WY
UT
50
30
r=0.58
30
r=0.52
2000
ME
CT
MI
VT
NJ
AKPA
MD
MO
MNOR
NH
WA
IA OH
COWI
DE
IL
NM
MT
CA
NV ND
WY
SD UTIN
KS
NE
AZ ID
60
1976
60
20
70
1964
CT NY
MD
NV
PACA
DE NM
NJ
MO
ILMI
HI MN
AK
WA
MT
WIOR
OH
NH
ID
UT
CO
WY
IN
ND
AZ
IA
ME
SD VT
NE KS
20
70
MN
RI MA
IA MD
PA
NY
WI
OR HI
IL WA
CA
VT
OH
MI
MO
DE
NM
NJ
CT
MT ME
IN
SD CO
ND
AZ
NVKS
NH AK
WYNE
ID
UT
r=0.16
20
70
1988
60
VT
30
70
20
ME
40
RI
MA
30
MA
40
70
60
40
r=0.12
1984
30
50
MO
60
CTMI
PA
WAMD
MO
NJ
OHAK
IL WI
CA
DE
OR
NH
NV
VT
MT
CO
SD
NMIA
IN
ND UT
AZ
KSWY
ID
NE
50
MN RI
MO
NM
AZ MT UT WA MA
CO
WI
WY
NV IA
ID OH
ILCANY
IN ORMICT
DE
MD
SD
NH PANJ
NE
NDKS
1960
70
MN
ME
20
40
20
1972
MA
60
NY
r=0.29
60
20
HI
30
50
30
60
70
UT
AZ
RI
WA
CA
MT
DE NV NM
MA
CT
OR
NH
ID MN
IL
MO MD
PA
NJ
WIMI
WY IAMEOH
IN
CO
ND
VT
SDNE
KS
70
1968
40
40
70
20
50
50
70
1956
60
50
60
r=0.12
20
70
1948
70
AZ UT
NVMT WA
MD
RI CA
NM
DE
IDOR CTPA
MA
WY NH
MO
OHWI
MN
NJ
NY
IL
MI
IN
IAMECO
VTND
NE
KS
SD
30
r=0.12
20
1944
HI
VT
RI
MDMANY
CA
DE
NJ
CT
IL ME
WA
OR
NM
MI
MN
WI
NV
CO PA
NHIA
OH
70
60
50
40
30
AZ
MOIN AKMT
NDSDKS NE
ID
WY
UT
r=0.89
20
2
−1
0
1
2
Policy Liberalism
Figure 5: The relationship between state policy liberalism and Democratic presidential vote
share, 1936–2012 (non-South only).
21
ticularistic spending. The sources of this discrepancy between the two measures—perhaps
di↵erences in political culture, budgetary decentralization, or economic need—could be an
interesting topic for future research.
1982
0.2
Policy Priorities (Jacoby & Schneider, 2009)
0.1
0.0
−0.1
−0.2
1990
AK
AZ
WY
NV
ID
ND UT NM DE
TX
WV WA
NC
IN
FL
AL
MT
SD
GA
KY
MS
AR SC
NECO IA OR
OK
VA
MN
KS
LA
TN
MD
VT
MO
NH ME
NJ
IL
WI
OH
CA
CT
PA
MI
RINY
MA
WY
NV
ID
AK
AZ NM
DE
NC
CO
ND
UTVA
MT
TX
MS
SD KS WV
AL
WAVT
GA FL KYMO
IAOR
AR SC LA OKIN
MD
TN NE
MN
NJ
WI
NHIL ME
CA
OH PA
RI
CT
MI
r=−0.57
NY
MA
r=−0.55
1998
2005
0.2
WY
UT
NV
ID
KS
0.1
AK
DEMT
SD OKIN
ND
VA CO NM WI
NC AZFL WVIAMI OR
GA
0.0
TX NE
WA MD
MO
VT NJ
MS AR
MN
KY
LA
OH
AL
TN
SC
CA
ILPA
−0.1
RI
MA
NH
CT
ME
−0.2
r=−0.44
−2
−1
1
NV
AK
MT
KS
ID
DE
UT
VT
SDVA CO
OK
TX IN WV MI WI
AR
OR
NJ
WA
NC
MD
FL
IA
GA
AZ
LA
NM MN
AL
KYNE
NH
CA
OH
SC
MO
IL
MS
CT
PA
MA
RI
ME
TN
r=−0.33
NY
0
WY
ND
2
−2
−1
NY
0
1
2
Policy Liberalism
Figure 6: The relationship between policy liberalism and policy priorities (Jacoby and Schneider 2009) in selected years, 1982–2005.
Dimensionality
Our one-dimensional model of state policies implies that a single latent trait captures systematic policy variation across states. This is not to say that it captures all policy di↵erences,
but it does imply that once policies’ characteristics and states’ policy liberalism are accounted for, any additional variation in state policies is essentially random. This assumption
would be violated if there were instead multiple dimensions of state policy, as some schol22
ars have claimed. Given that roll-call alignments in the U.S. Congress were substantially
two-dimensional for much of the 20th century (Poole and Rosenthal 2007), it is not unreasonable to suspect that state policies might be as well. As we demonstrate, however, a
one-dimensional model captures state policy variation surprisingly well, and there is little
value to increasing the complexity of the model by adding further dimensions.
1991
WA
Abortion Policy (NARAL, 2012)
NV
AK MT
CA
NMME
ORMDVT HI
CTNJ NY
r=0.79
10
WV
NHIL
CO
MA
DE IA MN
5
GA NC
MS
WY
TNAZ
WI
AR
SC
AL ND ID
SD
OK
UTVA
LA
FL
TX
KS
KY
MO INNE OH
−2
−1
RI
MI
PA
0
1
Environmental Innovation (Green Index, 1991−1992)
2011
CA
ME
WI
WAMI
OH
IL
20
LA
GA
SC
10
2
AL
Gay Rights (Lax and Phillips, 2009)
WA ME
NH
NM
HI
NY
RI
NV
0.50
MT
KY
AZ
0.25
GA
0.00
OR
CANJ
MS
MN
MD
DE
PA
ND SD
WY
OK
TN
KS
LA
FL IN
NE OH
WVAK
ALAR
SCNC UT
IDTX VA MO
−2
WI
−1
MI
0
1
2
Average Monthly AFDC Benefits Per Family (2010 dollars)
VT
CT
MA
CO
IA
PA
NH
HI
DE
AR
AK
−2
2008
ILIA
IN
MA
−1
0
1
2
Policy Liberalism
1.00
r=0.85
VACO
TX KS
CT NY
RI NJ
KY
NE
MO
TN
NM
UT
WY
AZ
ID OK
MT
ND SDNV
WV
Policy Liberalism
0.75
VT
MD
FL
NC
MS
OR
MN
r=0.72
30
Policy Liberalism
1988
CA
r=0.76
1000
MN
MI
WAWI
800
MT
ND
600
GA
NC
FL
400
AR
UT
VT
CT
OR
IA
KS
MD
CONENH
PAME
WYOK
WV
IL
OH
AZ
NV
DE
SD
VA ID MO
IN
KY
NM
NY
HI
MA
RI
NJ
SC
LATX
TN
200
MS
AL
−2
−1
0
1
2
Policy Liberalism
Figure 7: Relationships between policy liberalism and four issue-specific scales (abortion
rights, environmental protection, gay rights, and welfare benefits).
One fact in support for unidimensionality is that the most discriminating policies in our
dataset—those most strongly related to the latent factor—span a wide range of issues, including racial discrimination, women’s rights, gun control, labor law, energy policy, criminal
23
Table 1: Correlations between policy liberalism scales estimated using economic, social,
racial, and all policies. The unit of analysis is the state-year. The racial policy scale is
estimated for the 1950–70 period only.
All
Economic
Economic
0.92
Social
0.84
0.69
Racial
0.86
0.68
Social
0.55
rights, and welfare policy. Additional evidence is provided by the relationships between policy liberalism and four issue-specific scales: NARAL’s abortion rights scale (NARAL 2012),
the Green Index of Environmental Innovation in 1991–92 (Hall and Kerr 1991; Ringquist
and Garand 1999), a gay rights index derived from Lax and Phillips (2009b), and average
AFDC benefits per family in each state (Moffitt 2002). As Figure 7 shows, policy liberalism
accurately predicts variation within each of these disparate policy areas.
We can explore this question at a higher level of generality by scaling state policies within
each of three broad issue domains: economic, social, and racial.24 Policy cleavages in the
mass public and in the U.S. Congress are often considered to di↵er across these domains,
especially earlier in the 1936–2014 period (e.g., Layman, Carsey, and Horowitz 2006; Poole
and Rosenthal 2007). As the first column of the correlation matrix in Table 1 shows, however,
each domain-specific scale is strongly related to the policy liberalism scale based on all
policies. The domain-specific scales are also highly correlated with each other, with the
correlation being weakest for racial and social policies (estimated for 1950–70 only). On the
whole, Table 1 provides strong evidence that variation in state policies is one-dimensional
and does not vary importantly across issue domains.
As a final piece of evidence, we show that allowing for multiple latent dimensions does not
24. Because cross-state variation in civil rights policies is concentrated in the 1950–70 period, we estimate
the racial policy dimension for these two decades only.
24
substantially improve our ability to predict policy di↵erences between states. As our measure
of model fit we use percentage correctly predicted (PCP), which for binary variables is the
percentage of cases for which the observed value corresponds to its model-based predicted
value (0 or 1). In order to include ordinal and continuous variables in this calculation, we
convert them into binary variables by dichotomizing them at a threshold randomly generated
for each variable. We estimate one and two-dimensional probit IRT models separately in each
year using the R function ideal (Jackman 2012), which automatically calculates PCP. We
then evaluate how much the second dimension improves PCP (adding dimensions cannot
decrease PCP).
Based on this method, we find little evidence that adding dimensions improves our ability
to account for the data. In the average year, a one-dimensional model correctly classifies
82% of all dichotomized policy observations. Adding a second dimension increases average
PCP by only 1.5 percentage points. This improvement in model fit is less than the increase
in fit that is used in the congressional literature as a barometer of whether roll-call voting
in Congress has a one-dimensional structure (Poole and Rosenthal 2007, 33–4). Further,
the minimal improvement in model fit gained from adding a second dimension is consistent
across time—even during the mid-century heyday of two-dimensional voting in Congress.
Taken as a whole, the evidence supports two conclusions. First, a single latent dimension
captures the vast majority of policy variation across states across disparate policy domains.
This is true even at times when national politics was multidimensional. Second, the approximately 20% of cross-sectional policy variation not captured by a one-dimensional model does
not seem to have a systematic structure to it, or at least not one that can be described by
additional dimensions.
Substantive Applications
Our dynamic measure of policy liberalism opens up multiple avenues of research not possible
with cross-sectional measures. Most obviously, as we have shown, it permits descriptive
25
analyses of the ideological evolution of state policies over long periods of time. But the
availability of a dynamic measure also facilitates causal analyses that incorporate policy
liberalism as an outcome, treatment, or control variable. In particular, because it is available
for each state-year, our measure can be used in time-series–cross-sectional (TSCS) research
designs, which leverage variation across both units and time. The fact that our estimates
are available for nearly 80 years is especially valuable because TSCS estimators can perform
poorly unless the number of time units is large (e.g., Nickell 1981).
For example, scholars could examine how the cross-sectional relationship between state
public opinion and policy liberalism has evolved over time (Burstein 2003); estimate the statelevel relationship between changes in opinion and changes in policy (cf. Stimson, MacKuen,
and Erikson 1995); or analyze how interest groups or electoral institutions moderate the
opinion–policy link (cf. Gray et al. 2004; Lax and Phillips 2011). Or scholars could evaluate
the policy e↵ects of electoral outcomes or the partisan composition of state government
(cf. Erikson, Wright, and McIver 1989; T. Kousser 2002; Besley and Case 2003; Leigh 2008).
An alternative approach would be to analyze policy liberalism as a cause rather than an
e↵ect. For example, one prominent view is that citizens respond“thermostatically”to changes
in policy by moving in the ideologically opposite direction (Wlezien 1995). A related perspective argues that voters compensate for partisan e↵ects on policy through partisan balancing
(e.g., Erikson 1988; Alesina, Londregan, and Rosenthal 1993). Other scholars, however,
highlight the positive feedback e↵ects of policy changes (e.g., Pierson 1993; Campbell 2012).
Our policy liberalism estimates open up ways of adjudicating among these theories using
state-level TSCS designs.
The Policy E↵ects of Voter Registration Reforms
To illustrate the kinds of analyses made possible by our estimates, we conduct a brief investigation into the policy e↵ects of reforms designed to make voter registration easier. While
debate over such reforms often focuses on e↵ects on turnout or partisan advantage, their ef-
26
fects on policy are arguably most important.25 One intuitive theoretical prediction, derived
from median-voter models of redistribution, is that lowering registration barriers makes the
electorate larger and poorer, which in turn increases political support for redistributive (i.e.,
liberal) policies (Meltzer and Richard 1981; Husted and Kenny 1997).
The policy consequences of registration regulations specifically have been examined by
Besley and Case (2003, 35–7), who using a fixed-e↵ect (FE) framework find liberalizing
e↵ects of lower registration barriers on five state taxation and spending policies in the period
1958–98. Besley and Case’s two-way FE specification improves substantially over crosssectional comparisons, which cannot control for unobserved di↵erences between states. An
important weakness of their specification, however, is that it assumes that states did not
trend in di↵erent directions over the period they examine.26 Figure 2 suggests, however,
that this assumption is false (see, e.g., the liberalizing trend among Northeastern states).
The likely consequence is that Besley and Case’s e↵ect estimates are much too large.
We replicate and extend Besley and Case’s analysis, examining the policy e↵ects of three
electoral policies—“motor voter” laws, election-day registration, and mail-in registration—on
state policy liberalism between 1950 and 2000.27 To guard against di↵erential time trends,
we use a more conservative specification that includes a lagged dependent variable (LDV)
as well as state and year FEs.28 One advantage of a long time series is the finite-sample
bias of LDV-FE models is of order 1/T and thus decreases rapidly as the number of time
units increases (Beck and Katz 2011, 342). Table 2 reports the estimated e↵ect estimates,
all of which are positive and, except for motor voter registration, distinguishable from 0.
In terms of substantive magnitude, these estimates imply that making voter registration
easier increases the probability of a state adopting a liberal law by about 1 percentage point.
25. See, for example, Key’s (1949) and J. M. Kousser’s (1974) analyses of the policy e↵ects of su↵rage
restrictions in the post-Reconstruction South.
26. Besley and Case (2003) do include a few time-varying demographic controls, but these are unlikely to
account for di↵erential state trends.
27. We obtained data on the first two policies from Besley and Case (2003) and data on the third from
Springer (2014).
28. Following Besley and Case (2003), we define a unit-year as “treated” by a registration policy if that
policy was in e↵ect at the last election.
27
Consistent with our concern about state-specific trends, the estimates from a simple two-way
FE model (not shown) are all an order of magnitude larger than their LDV-FE counterparts.
Table 2: E↵ect of Electoral Reforms on State Policy Liberalism
Policy
Motor voter registration
0.012
(0.013)
Election day registration
0.035⇤⇤
(0.017)
Mail-in registration
0.021⇤⇤
(0.011)
Lagged Policy
0.925⇤⇤
(0.008)
Constant
0.007
(0.026)
FE for state
FE for year
X
X
Observations
R2
Adjusted R2
2,581
0.983
0.983
⇤
Note:
p<0.1;
⇤⇤
p<0.05
Though brief, this application highlights several advantages of our measure of policy liberalism. First, its TSCS structure enables us to exploit within-state variation in institutions
such as registration regulation. Second, its long time series permits the use of estimators,
such as LDV-FE models, whose performance improves as T increases. Third, the precision
of our composite measure relative to any individual indicator of liberalism means allows us
to detect small but meaningful e↵ects, such as the ones reported in Table 2.
28
Conclusion
This paper has addressed a major gap in the state politics literature: the lack of a measure
of state policy liberalism that varies across time. Using a dataset covering 148 policies and a
latent-variable model designed for a mix of ordinal and continuous data, we have generated
estimates of the policy liberalism of every state in every year for the past three-quarters
of a century. As indicated by their high correlations with existing measures of state policy
liberalism as well as with domain-specific indices, our estimates exhibit strong evidence of
validity as a measure of policy liberalism.
Our yearly estimates of policy liberalism are illuminating for their own sake, revealing
historical patterns in the development of state policymaking that would be hard to discern
otherwise. But they also open up research designs that leverage temporal variation in state
policies to explore questions involving the causes and e↵ects of policy outcomes. These
topics include the policy e↵ects of public mood, electoral outcomes, interest groups, and
institutions, as well as the consequences of policy change on political attitudes and behavior.
The relevance of this paper extends well beyond the field of state politics. In addition to
facilitating the study of topics of general significance, our measurement model could be applied to policymaking by local governments (cf. Tausanovitch and Warshaw 2014) as well as
in cross-national studies. Even more generally, our dynamic approach to measurement helps
to illustrate the value of data-rich, time-varying measures of important political concepts
like policy liberalism.
29
References
Adcock, Robert, and David Collier. 2001. “Measurement Validity: A Shared Standard for
Qualitative and Quantitative Research.” American Political Science Review 95 (3): 529–
546.
Alesina, Alberto, John Londregan, and Howard Rosenthal. 1993. “A Model of the Political
Economy of the United States.” American Political Science Review 87 (1): 12–33.
Ansolabehere, Stephen, Jonathan Rodden, and James M. Snyder Jr. 2008. “The Strength of
Issues: Using Multiple Measures to Gauge Preference Stability, Ideological Constraint,
and Issue Voting.” American Political Science Review 102 (2): 215–232.
Bafumi, Joseph, Andrew Gelman, David K. Park, and Noah Kaplan. 2005. “Practical Issues
in Implementing and Understanding Bayesian Ideal Point Estimation.” Political Analysis
13 (2): 171–187.
Beck, Nathaniel, and Jonathan N. Katz. 2011. “Modeling Dynamics in Time-Series–CrossSection Political Economy Data.” Annual Review of Political Science 14 (1): 331–352.
Berry, William D., Richard C. Fording, and Russell L. Hanson. 2000. “An Annual Cost of
Living Index for the American States, 1960–1995.” Journal of Politics 62 (2): 550–567.
Besley, Timothy, and Anne Case. 2003. “Political Institutions and Policy Choices: Evidence
from the United States.” Journal of Economic Literature 41 (1): 7–73.
Burstein, Paul. 2003. “The Impact of Public Opinion on Public Policy: A Review and an
Agenda.” Political Research Quarterly 56 (1): 29–40.
Campbell, Andrea Louise. 2012. “Policy Makes Mass Politics.” Annual Review of Political
Science 15:333–351.
Canak, William, and Berkeley Miller. 1990.“Gumbo Politics: Unions, Business, and Louisiana
Right-to-Work Legislation.” Industrial and Labor Relations Review 43 (2): 358–271.
30
Canes-Wrone, Brandice, David W. Brady, and John F. Cogan. 2002. “Out of Step, Out
of Office: Electoral Accountability and House Members’ Voting.” American Political
Science Review 96 (1): 127–140.
Chandler, Marsha, William Chandler, and David Vogler. 1974. “Policy Analysis and the
Search for Theory.” American Politics Research 2 (1): 107–118.
Clinton, Joshua, Simon Jackman, and Douglas Rivers. 2004. “The Statistical Analysis of Roll
Call Data.” American Political Science Review 98 (2): 355–370.
Converse, Philip E. 1964. “The Nature of Belief Systems in Mass Publics.” In Ideology and
Discontent, edited by David E. Apter, 206–261. London: Free Press.
Elazar, Daniel Judah. 1966. American Federalism: A View from the States. New York: Crowell.
Ellis, Christopher, and James A. Stimson. 2012. Ideology in America. New York: Cambridge
UP.
Erikson, Robert S. 1988. “The Puzzle of Midterm Loss.” Journal of Politics 50 (4): 1011–
1029.
Erikson, Robert S., Gerald C. Wright, and John P. McIver. 1989. “Political Parties, Public
Opinion, and State Policy in the United States.” American Political Science Review 83
(3): 729–750.
Fahrmeir, Ludwig, and Alexander Raach. 2007. “A Bayesian Semiparametric Latent Variable
Model for Mixed Responses.” Psychometrika 72 (3): 327–346.
Fiorina, Morris P., and Samuel J. Abrams. 2008. “Political Polarization in the American
Public.” Annual Review of Political Science 11 (1): 563–588.
31
Gamm, Gerald, and Thad Kousser. 2010. “Broad Bills or Particularistic Policy? Historical
Patterns in American State Legislatures.” American Political Science Review 104 (1):
151.
Gelman, Andrew. 2006. “Prior Distributions for Variance Parameters in Hierarchical Models.”
Bayesian Analysis 1 (3): 515–533.
Gray, Virginia, David Lowery, Matthew Fellowes, and Andrea McAtee. 2004. “Public Opinion, Public Policy, and Organized Interests in the American States.” Political Research
Quarterly 57 (3): 411–420.
Hall, Bob, and Mary Lee Kerr. 1991. 1991–1992 Green Index: A State-by-State Guide to the
Nation’s Environmental Health. Washington, DC: Island Press.
Ho↵erbert, Richard I. 1966. “The Relation between Public Policy and Some Structural and
Environmental Variables in the American States.” American Political Science Review
60 (1): 73–82.
Ho↵man, Matthew D., and Andrew Gelman. Forthcoming. “The No-U-Turn Sampler: Adaptively Setting Path Lengths in Hamiltonian Monte Carlo.” Journal of Machine Learning
Research.
Hopkins, Anne H., and Ronald E. Weber. 1976. “Dimensions of Public Policies in the American States.” Polity 8 (3): 475–489.
Husted, Thomas A., and Lawrence W. Kenny. 1997. “The E↵ect of the Expansion of the
Voting Franchise on the Size of Government.” Journal of Political Economy 105 (1):
54–82.
Jackman, Simon. 2009. Bayesian Analysis for the Social Sciences. Hoboken, NJ: Wiley.
32
Jackman, Simon. 2012. pscl: Classes and Methods for R Developed in the Political Science Computational Laboratory, Stanford University. Department of Political Science,
Stanford University. R package version 1.04.4. http://pscl.stanford.edu.
Jacoby, William G., and Saundra K. Schneider. 2001. “Variability in State Policy Priorities:
An Empirical Analysis.” Journal of Politics 63 (2): 544–568.
. 2009.“A New Measure of Policy Spending Priorities in the American States.”Political
Analysis 17 (1): 1–24.
. 2014. “State Policy and Democratic Representation.” In The Oxford Handbook of
State and Local Government, edited by Donald P. Haider-Markel. Oxford UP.
Key, V. O., Jr. 1949. Southern Politics in State and Nation. New York: Knopf.
King, Gary, Robert Keohane, and Sidney Verba. 1994. Designing Social Inquiry. Princeton,
NJ: Princeton UP.
Klingman, David, and William W. Lammers. 1984. “The ‘General Policy Liberalism’ Factor
in American State Politics.” American Journal of Political Science 28 (3): 598–610.
Kousser, J. Morgan. 1974. The Shaping of Southern Politics: Su↵rage Restriction and the
Establishment of the One-Party South. New Haven, CT: Yale University Press.
Kousser, Thad. 2002. “The Politics of Discretionary Medicaid Spending, 1980–1993.” Journal
of Health Politics, Policy and Law 27 (4): 639–672.
Krehbiel, Keith. 1998. Pivotal Politics: A Theory of U.S. Lawmaking. Chicago: University of
Chicago Press.
Ladd, Everett Carll, Jr. 1976. “Liberalism Upside Down: The Inversion of the New Deal
Order.” Political Science Quarterly 91 (4): 577–600.
Lax, Je↵rey R., and Justin H. Phillips. 2009a. “Gay Rights in the States: Public Opinion
and Policy Responsiveness.” American Political Science Review 103 (3): 367–386.
33
Lax, Je↵rey R., and Justin H. Phillips. 2009b. “How Should We Estimate Public Opinion in
The States?” American Journal of Political Science 53 (1): 107–121.
. 2011. “The Democratic Deficit in the States.” American Journal of Political Science
56 (1): 148–166.
Layman, Geo↵rey C., Thomas M. Carsey, and Juliana Menasce Horowitz. 2006. “Party Polarization in American Politics: Characteristics, Causes, and Consequences.” Annual
Review of Political Science 9 (1): 83–110.
Leigh, Andrew. 2008. “Estimating the Impact of Gubernatorial Partisanship on Policy Settings and Economic Outcomes: A Regression Discontinuity Approach.” European Journal of Political Economy 24 (1): 256–268.
Leiter, Richard A., ed. 2008. National Survey of State Laws. Detroit: Gale. Gale Virtual
Reference Library. Last accessed 11 January 2015.
Lowery, David, Virginia Gray, and Gregory Hager. 1989. “Public Opinion and Policy Change
in the American States.” American Politics Research 17 (1): 3–31.
Martin, Andrew D., and Kevin M. Quinn. 2002. “Dynamic Ideal Point Estimation via Markov
Chain Monte Carlo for the U.S. Supreme Court, 1953–1999.” Political Analysis 10 (2):
134–153.
Meltzer, Allan H., and Scott F. Richard. 1981. “A Rational Theory of the Size of Government.” Journal of Political Economy 89 (5): 914–927.
Moffitt, Robert. 2002. Welfare Benefits Data Base. Johns Hopkins University, Department
of Economics. http://www.econ2.jhu.edu/people/moffitt/datasets.html.
NARAL. 2012. Who Decides? The Status of Women’s Reproductive Rights in the United
States. http://www.prochoiceamerica.org/government-and-you/who-decides/.
34
Nickell, Stephen. 1981. “Biases in Dynamic Models with Fixed E↵ects.” Econometrica 49 (6):
1417–1426.
Noel, Hans. 2014. Political Ideologies and Political Parties in America. New York: Cambridge
University Press.
Persson, Torsten, and Guido Tabellini. 2006. “Electoral Systems and Economic Policy.” In
The Oxford Handbook of Political Economy, 725–738. New York: Oxford University
Press.
Pierson, Paul. 1993. “When E↵ect Becomes Cause: Policy Feedback and Political Change.”
World Politics 45 (4): 595–628.
Poole, Keith T., and Howard Rosenthal. 2007. Ideology & Congress. New Brunswick, NJ:
Transaction Publishers.
Quinn, Kevin M. 2004. “Bayesian Factor Analysis for Mixed Ordinal and Continuous Responses.” Political Analysis 12 (4): 338–353.
R Core Team. 2013. R: A Language and Environment for Statistical Computing. Vienna,
Austria: R Foundation for Statistical Computing, http://www.R-project.org/.
Ringquist, Evan J., and James C. Garand. 1999. “Policy Change in the American States.” In
American State and Local Politics: Directions for the 21st Century, edited by Ronald E.
Weber and Paul Brace, 268–99. New York: Chatham House: Seven Bridges Press.
Schickler, Eric. 2013. “New Deal Liberalism and Racial Liberalism in the Mass Public,
1937–1968.” Perspectives on Politics 11 (1): 75–98.
Sharkansky, Ira, and Richard I. Ho↵erbert. 1969. “Dimensions of State Politics, Economics,
and Public Policy.” American Political Science Review 63 (3): 867–879.
Shor, Boris, and Nolan McCarty. 2011. “The Ideological Mapping of American Legislatures.”
American Political Science Review 105 (3): 530–51.
35
Sorens, Jason, Fait Muedini, and William P. Ruger. 2008. “US State and Local Public Policies
in 2006: A New Database.” State Politics & Policy Quarterly 8 (3): 309–326.
Springer, Melanie Jean. 2014. How the States Shaped the Nation: American Electoral Institutions and Voter Turnout, 1920–2000. Chicago: University of Chicago Press.
Stan Development Team. 2013. Stan: A C++ Library for Probability and Sampling, Version
1.3. http://mc-stan.org/.
Stimson, James A. 1991. Public Opinion in America: Moods, Cycles, and Swings. Boulder,
CO: Westview.
Stimson, James A., Michael B. MacKuen, and Robert S. Erikson. 1995. “Dynamic Representation.” American Political Science Review 89 (3): 543–565.
Tausanovitch, Chris, and Christopher Warshaw. 2014. “Representation in Municipal Government.” American Political Science Review 108 (3): 605–641.
Walker, Jack L. 1969. “The Di↵usion of Innovations among the American States.” American
Political Science Review 63 (3): 880–899.
West, Mike, and Je↵ Harrison. 1997. Bayesian Forecasting and Dynamic Models. New York:
Springer.
Wlezien, Christopher. 1995. “The Public as Thermostat: Dynamics of Preferences for Spending.” American Journal of Political Science 39 (4): 981–1000.
Wright, Gerald C., Robert S. Erikson, and John P. McIver. 1987. “Public Opinion and Policy
Liberalism in the American States.” American Journal of Political Science 31 (4): 980–
1001.
36
Online Appendix: “The Dynamics of State Policy Liberalism,
1936–2012”
Table 1: Description of Policies
Policy
Years
Description
Sources
Abortion Policies:
Access for Contraceptives
1974-2014
Can pharmacies dispense emergency contraception without a prescription?
Does the state mandate counseling before an abortion?
Does the state mandate counseling before an abortion?
Did the state allow abortion before Roe v. Wade?
Does the state require parental notification or consent
prior to a minor obtaining an abortion?
Does the state ban late-term or partial birth abortions?
Does the state’s Medicaid system pay for abortions?
[78, 100, 106, 104]
Forced Counseling before Abortions
Forced Counseling before Abortions
Legal Abortion Pre-Roe
Parental Notification/Consent Required for
Abortion
Partial Birth Abortion Ban
Medicaid for Abortion
Criminal Justice Policies:
Age Span Provisions for Statutory Rape
1992-2014
1973-1991
1967-1973
1983-2014
1996-2000
1981-2014
1950-1998
Death Penalty
Drug & Alcohol Policies:
Beer Keg Registration Requirement
1936-2014
Decriminalization of Marijuana Possession
Medical Marijuana
Minimum Legal Drinking Age 21
Smoking Ban - Workplaces
Smoking Ban - Restaurants
Zero Tolerance (<.02 BAC) for Underage
Drinking
Education Policies:
Allow Ten Commandments in Schools
1973-2014
1996-2014
1936-1985
1995-2014
1995-2014
1983-1995
Ban on Corporal Punishment in Schools
Education Spending Per Pupil
1970-2014
1936-2009
Moment of Silence Required
1957-2014
Per Student Spending on Higher Education
Teacher Degree Required - High School
1988-2013
1936-1963
Teacher Degree Required - Elementary
1936-1969
School for Deaf
1936-1950
State Library System
Compulsory Education - Age
Environmental Policies:
Air Pollution Control Acts (Pre-CAA)
1936-1955
1936-1939
1947-1967
Bottle Bill
1970-2014
CA Car Emissions Standard
2003-2012
Electronic Waste Recycling Program
2000-2014
Endangered Species Act
Environmental Protection Act
1969-2014
1969-2014
Greenhouse Gas Cap
2006-2014
1978-2013
1936-2013
[78,
[78,
[84,
[78,
100, 136]
100, 136]
48]
100, 99, 52]
[78, 100, 8, 51]
[78, 100, 53, 8, 96]
Does a state adopt an age span provision into its statutory
rape law which e↵ectively decriminalizes sexual activity
between similar-aged teens?
Has the state abolished the death penalty?
[26]
Does the state require the registration upon purchase of a
beer keg?
Is marijuana possession a criminal act?
Is it legal to use marijuana for medical purposes?
Does the state have a minimum legal drinking age of 21?
Does the state ban smoking in all workplaces?
Does the state ban smoking in restaurants?
Does the state have a Zero Tolerance law for blood alcohol
levels <0.02 for individuals under age 21?
[78, 120, 176]
Does the state allow the Ten Commandments to be posted
in educational institutions?
Does the state ban corporal punishment in schools?
What is the per capita spending on public education per
pupil based on daily average attendance?
Does the state have a mandatory moment of silence period
at the beginning of each school day?
What is the per student subsidy for higher education?
In what year does the state require high school teachers
to hold a degree?
In what year does the state require elementary school
teachers to hold a degree?
In what year did the state establish residential schools for
the deaf?
In what year did the state establish a library system?
At what age are children allowed to leave school?
Does the state have an air pollution control act (Pre-Clean
Air Act)?
Does the state require a deposit on bottles paid by the
consumer and refunded when the consumer recycles?
Does the state adopt California’s car emissions standards
(which are more stringent than the federal level)?
Does the state have a recycling program for electronic
waste?
Does the state have an endangered species act?
Does the state have its own version of the federal National
Environmental Policy Act?
Does the state have a binding cap on greenhouse gas emissions in the utility sector?
1
[32]
[127, 86]
[91, 109]
[119]
[11, 25]
[11, 25]
[119]
[35, 5]
[78, 50]
[151]
[35, 108, 79]
[144]
[112]
[112]
[163]
[162]
[154]
[6, 102]
[29]
[92]
[37, 146, 36]
[78, 146, 13]
[78, 90, 178]
[134, 19, 23]
Description of Policies – Continued from previous page
Policy
Years
Description
Sources
Public Benefit Fund
1996-2014
[24, 118, 31]
Solar Tax Credit
1975-2014
Does the state have a public benefit fund for renewable
energy and energy efficiency?
Does the state have a tax credit for residential solar installations?
1977-2012
1964-2014
Does the state allow casinos?
Does the state have a lottery?
[10]
[132, 97]
1989-2014
Does the state ban discrimination against gays in public
accommodations?
Does the state allow civil unions or gay marriage (ordinal)?
Does the state forbid employment discrimination on the
basis of sexual orientation and/or sexual identity?
Are hate crimes explicity illegal in the state?
Does the state forbid sodomy?
[58]
Gambling Policies:
Casinos Allowed
Lottery Allowed
Gay Rights Policies:
Ban on Discrimination - Public Accommodation
Civil Unions and Gay Marriage
Employment Discrimination Protections
2000-2012
1982-2014
Hate Crimes Ban
Sodomy Ban
Gun Control Policies:
Assault Weapon Ban
Background Check - Dealer Purchase
1999-2014
1962-2003
1989-2014
1936-1993
Background Check - Private Sales
1936-2014
Gun Dealer Licenses
1936-2014
Gun Purchases - Waiting Period
Open Carry Law
Saturday Night Special
“Stand Your Ground” Law
Gun Registration
Immigration Policies:
English as Official Language
Instate Tuition for Immigrants
Labor Rights Policies:
Age discrimination ban
Anti-Injunction Act
Collective Bargaining - State Employees
Collective Bargaining - Teachers
1960-1996
Disability Discrimination Ban
Merit System for State Employees
Minimum Wage above Federal Level
[78, 121, 177]
[78, 115, 110]
[78, 115, 116]
[78, 58, 131]
[43, 9]
[78, 68, 175]
[78, 175, 69]
1936-2014
1961-2014
1974-2013
1993-2014
1936-2014
Are assault weapons banned in the state?
Does the state require a background check on gun purchases from dealers?
Does the state require a background check on privately
sold guns?
Does the state have any license requirements for manufacturers or dealers?
Does the state have a waiting period for gun purchases?
Is there an open carry law for guns?
Does the state ban “Saturday Night Special” handguns?
Does the state have a “stand your ground” law?
Does the state have a registration requirement for guns?
1961-2014
2001-2014
Is English the state’s official language?
Does the state allow in-state tuition for illegal immigrants?
[38]
[105]
1936-1999
1936-1966
1966-1996
[124, 123, 67]
[170]
[46, 27, 174]
1965-1990
1936-1953
1968-2012
Does the state ban age discrimination in hiring?
Does the state have an anti-injunction law?
Does the state have collective bargaining rights for state
government employees?
Does the state have collective bargaining rights for local
teachers?
Does the state ban discrimination against disabled people?
Does the state have a merit system for state employees?
Is the state’s minimum wage above the federal level?
Minimum Wage for Men
1944-1968
Does the state have a minimum wage for men?
Minimum Wage for Women
1936-1980
Does the state have a minimum wage for women?
Prevailing Wage Law
Right to Work Law
State Pension System
Temporary Disability Insurance
1936-2014
1944-2014
1936-1960
1945-2014
Unemployment Compensation
1937-2014
Workers’ Compensation
Child Labor Work Certificates
1936-1947
1936-1939
Labor Relations Acts
1937-1966
Does the state have prevailing wage laws?
Is the state a right-to-work state?
When did the state establish its pension system?
Does the state have a temporary disability insurance program?
What is the maximum weekly amount of unemployment
benefits?
Has the state established workers’ compensation?
Does the state require employment certificates for child
labor (14 and 15)?
Does the state have a Labor Relations Act?
Licensing Policies:
Chiropractor Licensing
Dentist Licensing
Architect Licening
Beautician Licensing
Pharmacist Licensing
Engineer Licensing
1936-1951
1936-1951
1936-1951
1936-1951
1936-1951
1936-1951
When
When
When
When
When
When
did
did
did
did
did
did
2
the
the
the
the
the
the
state
state
state
state
state
state
require
require
require
require
require
require
licensing
licensing
licensing
licensing
licensing
licensing
for
for
for
for
for
for
chiropractors?
dentists?
architects?
beauticians?
pharmacists?
engineers?
[78, 175, 74]
[78, 175, 70]
[78,
[78,
[78,
[78,
[78,
175,
175]
175,
175,
175,
75]
71]
73]
72]
[34, 113, 114, 80,
46, 27, 174]
[62]
[148]
[139, 140, 141, 166,
167, 168, 171]
[139, 140, 141, 166,
167, 168, 171]
[139, 140, 141, 166,
167, 168, 171]
[157, 66]
[4, 28, 85]
[149]
[158]
[137, 165]
[41, 169, 41]
[93, 173, 94, 95,
161, 81, 88, 89, 83]
[103, 164, 172, 88,
89, 83]
[150]
[150]
[150]
[150]
[150]
[150]
Description of Policies – Continued from previous page
Policy
Years
Description
Sources
Nurse Licensing
Accountant Licensing
Real Estate Licensing
Misc. Regulatatory Policies:
Anti-sedition Laws
Compulsory Sterilization
1936-1951
1936-1951
1936-1951
When did the state require licensing for nurses?
When did the state require licensing for accountants?
When did the state require licensing for real estate agents?
[150]
[150]
[150]
1936-1955
1945-1974
[1]
[63]
Grandparents’ Visitation Rights
1964-1987
Hate Crimes Ban
Urban Housing - Enabling Federal Aid
Urban Housing - Direct State Aid
Living Wills
1981-2014
1936-1953
1939-1951
1976-1992
Pain and Su↵ering Limits in Lawsuits
1961-2012
Physician-assisted suicide
Planning Laws Required for Local Gov.
1998-2014
1961-2007
Protections Against Compelling Reporters to
Disclose Sources
Rent Control Prohibition
1936-2013
Religious Freedom Restoration Act
State Debt Limitation
Municipal Home Rule
1993-2014
1936-1966
1936-1961
Lemon Laws
1970-2014
Utility Regulation
Cruelty to Animals
1936-1960
1936-2014
Does the state have anti-sedition laws?
Does the state have a forced sterilization program (directed toward the disabled, delinquent, etc.)?
Does the state have a law guaranteeing grandparents’ visitation rights?
Are hate crimes explicity illegal in the state?
Does the state have a law enabling federal housing aid?
Does the state provide direct aid for urban housing?
Does the state have a law permitting individuals control
over the use of heroic medical treatment in the event of a
terminal illness?
Are there limits on damages for pain and su↵ering in lawsuits?
Does the state allow physician-assisted suicide?
Does the state have a law authorizing or requiring growthmanagement planning?
Does the state have a Shield Law protecting journalists
from revealing their sources?
Does the state prohibit the passage of rent control laws in
its cities or municipalities?
Did the state pass the Religious Freedom Restoration Act?
In what year did the state establish debt limitation?
Did the state pass a law enabling voters to adopt a municipal home rule charter?
Did the state pass a law protecting consumers who purchase automobiles which fail after repeated repairs?
In what year did the state regulate utilities?
Has the state made aggravated animal cruelty a first- or
second-o↵ense felony?
Racial Discrimination Policies:
School Segregation
Ban on Interracial Marriage
Banning discrimination in public accommodations (pre-CRA)
1936-1953
1936-1967
1936-1963
Did the state require segregation in public schools?
Does the state have a law banning interracial marriages?
Does the state pass a law (with adminstrative enforcement) banning discrimination in public accomodations
(pre-Civil Rights Act)?
Does the state pass a law (with adminstrative enforcement) banning discrimination in public accomodations
(post-Civil Rights Act)?
Does the state have a fair employment law?
Does the state have a fair employment law? (post-1964)
Does the state ban discrimination in private housing?
Does the state ban discrimination in public housing?
Does the state have urban renewal areas?
Does the state have a cigarette tax?
What is the state’s tax on a pack of cigarettes?
Does the state have an earned income tax credit?
Does the state have an income tax?
What is the state individual income tax rate for an individual that makes more than 1.5 million real dollars?
Does the state have a sales tax?
What is the sales tax rate?
What is the state’s tax burden (per capita taxes/per capita
income)?
Is there a corporate income tax?
What is the highest corporate tax rate?
Is there a state estate tax?
[78, 98, 133, 33, 16]
[78, 98, 133, 33, 16]
[78, 82, 22]
In what year did the state enact controlled-access highways?
Does the state require that people use helmets while on
bicycles?
[77, 20]
1950-2014
Banning discrimination in public accommodations (post-CRA)
1964-2010
Fair Employment Laws
Fair Employment Laws (post-1964)
Fair Housing - Private Housing
Fair Housing - Public Housing
Fair Housing - Urban Renewal Areas
Cigarette Tax
Cigarette Tax Rate
Earned Income Tax Credit
Income Tax
Income Tax Rate - Wealthy
1945-1964
1965-2014
1959-1968
1937-1964
1945-1964
1936-1946
1947-2014
1988-2014
1936-2014
1977-2012
Sales Tax
Sales Tax Rate
Tax Burden
1936-1945
1946-2014
1977-2010
Corporate Income Tax
Top Corporate Tax Rate
Estate Tax
Transportation Policies:
Controlled Access Highways
1936-1940
1941-2014
2009-2014
Bicycle Helmets Required
1985-2014
1937-1946
3
[45, 57]
[60]
[59]
[30]
[55]
[78, 160]
[78, 130, 126]
[128]
[78, 135, 129]
[78, 101, 61]
[54]
[56]
[87]
[78, 3, 76, 138]
[145]
[78, 14, 12, 42]
[78, 82, 22]
[16, 156, 39]
[16, 156, 39]
[16, 156, 39]
[16, 156, 39]
[59]
[151, 147]
[151, 147]
[151, 147]
[151, 147]
[151, 147]
[151, 147]
[151, 147]
[151, 147]
[151, 147]
[151, 147]
[151, 147]
[18]
Description of Policies – Continued from previous page
Policy
Years
Description
Sources
Mandatory Seat Belts
1984-2014
[40]
Motorcycle Helmets Required
1967-2014
Mandatory Car Insurance
Welfare Policies:
AFDC - Benefits for Average Family
1945-2012
Does the state require the usage of seat belts (either primary or secondary enforcement)?
Does the state require the usage of helments by people on
motorcycles?
Does the state require drivers to obtain car insurance?
AFDC-UP Policy
1961-1990
Aid to Blind - Average Payment per Recipient
1936-1965
Aid to Blind - Avgerage Payment per Recipient (post-1965)
Aid to Disabled - Average Payment per Recipient
Aid to Disabled - Average Payment per Recipient (post-1965)
CHIP - Eligibility Level for Children
CHIP - Eligibility Level for Infants
CHIP - Eligibility Level for Pregnant Women
General Assistance Payment Per Case
1966-1972
General Assistance Payment Per Recipient
1964-1980
Old Age Assistance - Average Payment per Recipient
Old Age Assistance - Average Payment per Recipient (post-1965)
Senior Prescription Drugs
1936-1965
State Adoption of Medicaid
Medicaid - Eligibility Level for Pregnant
Women
TANF - Average Payment per Family
1966-1983
1990-1997
TANF - Initial Eligibility Level
1996-2013
TANF - Max Payments
1990-2013
Women’s Rights Policies:
Equal Pay
1936-1972
ERA Ratification
State Equal Rights Law
1972-2014
1971-2014
Gender Discrimination Laws
1961-1964
Gender Discrimination Laws (post-1964)
1965-2014
No Fault Divorce
Jury Service for Women
1966-2014
1936-1967
1936-1992
1951-1965
1966-1972
1998-2012
1998-2012
1998-2012
1937-1963
1966-1972
1975-2001
2006-2010
[47]
[159]
What is the average level of benefits per family under the
Aid for Families with Dependent Children program?
What is the average level of benefits under the Aid for
Families with Dependent Children program?
What is the average monthly payment per recipient for
the permanently blind or disabled?
What is the average monthly payment per recipient for
the permanently blind or disabled? (post-1965)
What is the average monthly payment per recipient for
the permanently blind or disabled?
What is the average monthly payment per recipient for
the permanently blind or disabled? (post-1965)
What is the CHIP eligibility level for children?
What is the CHIP eligibility level for infants?
What is the CHIP eligibility level for pregnant women?
What is the average monthly payment per case for general
assistance (an early form of welfare)?
What is the average monthly payment per recipient for
general assistance (an early form of welfare)?
What is the average monthly payment per recipient per
recipient for old age assistance?
What is the average monthly payment per recipient per
recipient for old age assistance? (post-1965)
Does the state provide pharmaceutical coverage or assistance for seniors who do not qualify for Medicaid?
Does the state have a Medicaid program?
What is the Medicaid eligibility level for pregnant women?
[17, 142, 143, 155]
What is the average monthly level of benefits per family
under the Temporary Aid for Needy Families program?
What is the initial eligibility level for benefits for a family of three under the Temporary Aid for Needy Families
Program?
What is the maximum level of benefis under the Temporary Aid for Needy Families program for a family of three
with no income?
[152]
Does the state have a law providing for equal pay for
women working in the same job?
Has the state ratified the Equal Rights Amendment?
Has the state passed a state-level equivalent to the Equal
Rights Amendment?
Does the state ban hiring discrimination on the basis of
gender?
Does the state ban hiring discrimination on the basis of
gender? (post-1964)
Does the state have a no-fault divorce policy?
Can women serve on juries?
[78, 107, 2, 21]
4
[15]
[17, 142, 143, 155]
[142, 143, 155]
[142, 143, 155]
[142, 143, 155]
[117]
[117]
[117]
[17, 142, 143, 155]
[142, 143, 155]
[17, 142, 143, 155]
[142, 143, 155]
[111, 65]
[49]
[117]
[152]
[152]
[7, 122, 64]
[78, 44, 153]
[125]
[125]
[151]
[151]
References
[1] State sedition laws: Their scope and misapplication. Indiana Law Journal, 31(2):270–285, 1956.
[2] Equal Pay for Equal Work. The Bureau of National A↵airs, Washington, D.C., 1963.
[3] A sour note: A look at the Minnesota lemon law. Minnesota Law Review, 68(1):846–880, 1983.
[4] The new encyclopedia of Southern culture. In Griffin, Larry J., Peggy G. Hargis, and Charles Reagan Wilson, editor,
Social Class, volume 20. University of North Carolina Press, Chapel Hill, NC, 2012.
[5] Abdel-Monem, Tarik. Posting the Ten Commandments as a historical document in public schools. Iowa Law Review,
87:1023–1057, January 2000.
[6] Aborn, Richard A. and Carl E. Axelrod. State air pollution control legislation. Boston College Law Review, 9:712–756,
April 1968.
[7] Alice Paul Institute/National Council of Women’s Organizations. The Equal Rights Amendment: Unfinished business
for the Constitution. http://www.equalrightsamendment.org/states.htm#AR.
[8] American Civil Liberties Union. Public funding for abortion.
https://www.aclu.org/reproductive-freedom/public-funding-abortion, July 2004.
[9] American Civil Liberties Union (ACLU). History of sodomy laws and the strategy that led to today’s decision.
https://www.aclu.org/lgbt-rights_hiv-aids/history-sodomy-laws-and-strategy-led-todays-decision, June
2003.
[10] American Gaming Association. State of the states: The AGA survey of casino entertainment.
http://www.americangaming.org/industry-resources/research/state-states, 1999–2014.
[11] American Nonsmokers’ Rights Foundation (ANRF). Chronological table of U.S. population protection by 100%
smokefree state or local laws. http://www.no-smoke.org/pdf/EffectivePopulationList.pdf, January 2015.
[12] American Prosecutors Research Institute. Animal cruelty prosecution: Opportunities for early response to crime and
interpersonal violence. http://www.ndaa.org/pdf/animal_cruelty_06.pdf, 2006.
[13] Animal Legal and Historical Center. Statutes. https://www.animallaw.info/statutes.
[14] Animal Legal Defense Fund. U.s. jurisdictions with and without felony animal provisons. http://aldf.org/resources/
advocating-for-animals/u-s-jurisdictions-with-and-without-felony-animal-cruelty-provisions/.
[15] Anne E. Winkler. Does AFDC-UP encourage two-parent families? Journal of Policy Analysis and Management,
14(1):4–24, 1995.
[16] Barnett, Richard and Joseph Garai. Where the States Stand on Civil Rights. Sterling Publishing Co, Inc., New York,
1962.
[17] Bateman, Mary E. Trends in public assistance, 1933-1939. Division of Public Assistance Research, Washington, D.C.,
1940.
[18] Bicycle Helmet Safety Institute. Helmet laws for bicycle riders. http://www.helmets.org/mandator.htm.
[19] Bifera, Lucas. Regional greenhouse gas initiative.
http://www.c2es.org/docUploads/rggi-brief-12-18-13-updated.pdf, December 2013.
[20] Breed, C.B., Cli↵ord Older, and W.S. Downs. Highway costs: A study of highway costs and motor vehicle payments in
the United States. Submitted to Association of American Railroads, January 1939.
[21] John Marshall Butler. The case for equal legal rights. National Business Woman, XXXVI:6–16, 1957.
[22] Caldwell, Wallace F. State public accommodation laws, fundamental liberties and enforcement programs. Washington
Law Review, 40:841–872, 1965.
[23] California Environmental Protection Agency. Assembly Bill 32 overview. http://www.arb.ca.gov/cc/ab32/ab32.htm.
[24] Center for Climate and Energy Solutions. Public benefit funds.
http://www.c2es.org/us-states-regions/policy-maps/public-benefit-funds, 2015.
5
[25] Centers for Disease Control and Prevention (CDC). State smoke-free laws for worksites, restaurants, and bars – United
States, 2000-2010. Morbidity and Morality Weekly Report, 60(15):472–475, April 2011.
[26] Cocca, Carolyn E. The politics of statutory rape laws: Adoption and reinvention of morality policy in the states,
1971-1999. Polity, 35(1):51–72, 2002.
[27] Cohen, Emily, Kate Walsh, and RiShawn Biddle. Invisible ink in collective bargaining: Why key issues are not
addressed, July 2008.
[28] Collins, Benjamin. Right to work laws: Legislative background and empirical research, January 2014.
[29] Container Recycling Initiative. Bottle bill resource guide: All U.S. bottle bills.
http://www.bottlebill.org/legislation/usa/allstates.htm, June 2009.
[30] Council of State Governments. The Book of the States, page 381. 1952.
[31] Database of State Incentives for Renewables & Efficiency (DSIRE). Public benefits funds.
http://www.dsireusa.org/solar/solarpolicyguide/?id=22.
[32] Death Penalty Information Center. States with and without death penalty.
http://www.deathpenaltyinfo.org/states-and-without-death-penalty.
[33] Douglas, Davidson M. The limits of the law in accomplishing racial change: School segregation in the pre-Brown North.
Faculty Publications, Journal of William and Mary Law School, Paper 118, 1997.
[34] Education Commission of the States. ECS state notes & policy, 2008.
[35] Education Commission of the States (ECS). School prayer, moment of silence, other policies concerning religion.
http://www.ecs.org/clearinghouse/13/38/1338.htm, July 2000.
[36] Electronics Recycling Coordination Clearinghouse. Map of states with legislation.
http://www.ecycleclearinghouse.org/content.aspx?pageid=10.
[37] Electronics TakeBack Coalition. Brief comparison of state laws on electronics recycling.
http://www.electronicstakeback.com/wp-content/uploads/Compare_state_laws_chart.pdf, September 2013.
[38] English Language Association. Official english map. https://www.proenglish.org/official-english/state-profiles.
[39] Estrep, Samuel D. Current Trends in State Legislation. University of Michigan Law School, Ann Arbor, MI, 1952.
[40] Farmer, Charles M. and Allan F. Williams. E↵ect of fatality risk of changing from secondary to primary seatbelt
enforcement. http://www.gahighwaysafety.org/pdf/iihsseatbeltsof.pdf, December 2004.
[41] Fishback, Price V. and Shawn Everett Kantor. The adoption of worker’s compensation in the United States, 1900-1930.
Journal of Law and Economics, 41(2):305–342, October 1998.
[42] Frasso, Paula J. The Massachusetts animal cruelty statute: A real dog – a proposal for a redraft of the current law.
New England Law Review, 35:1003, 2000–2001.
[43] Gay and Lesbian Archives of the Pacific Northwest (GLAPN). Sodomy laws in the united states.
http://www.glapn.org/sodomylaws/usa/usa.htm, November 2007.
[44] Gladstone, Leslie W. Equal Rights Amendment: State provisions, August 2004.
[45] Goldberg, Michael K. A survey of the fifty states’ grandparent visitation statutes. Marquette’s Elder Advisor,
10(2):245–268, 2009.
[46] Goldfield, Michael and Amy Bromsen. The changing landscape of U.S. unions in historical and theoretical perspective.
Annual Review of Political Science, 16:231–257, 2013.
[47] Governors Highway Safety Association. Helmet laws – motorcycle helmets.
http://www.ghsa.org/html/stateinfo/laws/helmet_laws.html, February 2015.
[48] Greenhouse, Linda and Reva Siegel. Before Roe v. Wade: Voices that shaped the abortion debate before the Supreme
Court’s ruling. Yale Law School, Public Law Working Paper, 257, 2012.
[49] Gruber, Jonathan. Medicaid. http://www.nber.org/papers/w7829.pdf, August 2000.
6
[50] Gunderson National Child Protection Training Center. Discipline and the law: State laws.
http://www.gundersenhealth.org/ncptc/center-for-effective-discipline/discipline-and-the-law/state-laws.
[51] Guttmacher Institute. State policies in brief: Bans on ‘partial-birth’ abortion.
http://www.guttmacher.org/statecenter/spibs/spib_BPBA.pdf, November 2014.
[52] Guttmacher Institute. State policies in brief: Parental funding of minor’s abortions.
http://www.guttmacher.org/statecenter/spibs/spib_PIMA.pdf, June 2014.
[53] Guttmacher Institute. State policies in brief: State funding of abortion under Medicaid.
http://www.guttmacher.org/statecenter/spibs/spib_SFAM.pdf, January 2015.
[54] Hamilton, Marci A. Development of state RFRA statutes. http://rfraperils.com/states/, December 2014.
[55] Hays, Scott P. Patterns of reinvention: the nature of evolution during policy di↵usion. Policy Studies Journal,
24(4):551–566, 1996.
[56] Heins, A. James. Constitutional Restrictions Against State Debt. University of Wisconsin Press, Madison, WI, 1963.
[57] Hill, Twyla J. Legally extending the family: An event history analysis of grandparent visitation rights laws. Journal of
Family Issues, 21(2):246–261, 2000.
[58] Human Rights Campaign. Maps of state laws and policies. http://www.hrc.org/state_maps.
[59] Intergroup Relations Service. Fair Housing Laws. U.S. Housing and Home Finance Agency, Washington, D.C., 1965.
[60] Jenness, Valerie and Ryken Grattet. Making Hate a Crime: From Social Movement to Law Enforcement. Russell Sage
Foundation, New York, 2004.
[61] John Willis. State rent-control legislation, 1946-1947. Yale Law Journal, 57:351–376, 1947.
[62] Jolls, Christine and J.J. Prescott. Disaggregating employment protection: The case of disability discrimination. Harvard
Law School John M. Olin Center for Law, Economics, and Business Discussion Paper Series, 496, September 2004.
[63] Kaelber, Lutz. Eugenics: Compulsory sterilization in 50 American states (database).
http://www.uvm.edu/~lkaelber/eugenics/.
[64] Kanowitz, Leo. The ERA: The task ahead. Hastings Constitutional Law Quarterly, 6:637–662, 1979.
[65] Karch, Andrew. Democratic Laboratories: Policy Di↵usion Among the American States. University of Michigan Press,
Ann Arbor, MI, 2007.
[66] Kelsay, Michael P., James I. Sturgeon, and Kelly D. Pinkham. The adverse economic from repeal of the prevailing wage
law in Missouri. http://cas.umkc.edu/economics/resources/prevailingwagestudy.pdf, December 2011.
[67] Lahey, Joanna. State age protection laws and the age discrimination in employment act. National Bureau of Economic
Research, Working Paper 12048, February 2006.
[68] Law Center to Prevent Gun Violence. Assault weapons policy summary.
http://smartgunlaws.org/assault-weapons-policy-summary/#state, June 2013.
[69] Law Center to Prevent Gun Violence. Background check procedures policy summary.
http://smartgunlaws.org/background-check-procedures-policy-summary/, November 2013.
[70] Law Center to Prevent Gun Violence. Dealer regulations policy summary.
http://smartgunlaws.org/dealer-regulations-policy-summary/, September 2013.
[71] Law Center to Prevent Gun Violence. Design safety standards policy summary.
http://smartgunlaws.org/gun-design-safety-standards-policy-summary/, December 2013.
[72] Law Center to Prevent Gun Violence. Registration of firearms summary.
http://smartgunlaws.org/registration-of-firearms-policy-summary/, October 2013.
[73] Law Center to Prevent Gun Violence. “Stand Your Ground” policy summary.
http://smartgunlaws.org/stand-your-ground-policy-summary/, July 2013.
7
[74] Law Center to Prevent Gun Violence. Universal background checks & the private sale loophole policy summary.
http://smartgunlaws.org/universal-gun-background-checks-policy-summary/, August 2013.
[75] Law Center to Prevent Gun Violence. Waiting periods policy summary.
http://smartgunlaws.org/waiting-periods-policy-summary/, June 2013.
[76] Leiter, Richard A. National Survey of State Laws. Thomas Gale, Detroit, 2008.
[77] Levin, David R. Public Control of Highway Access and Roadside Development. United States Government Printing
Office, Washington, D.C., 1947.
[78] Lexis Nexis/State Governments. State statutes.
[79] Liberty Council. State laws regarding moment of silence in public schools.
http://www.lc.org/media/9980/attachments/moment_of_silence_pledge_of_allegiance_state_listing_061702.pdf,
2002.
[80] Lindy, Benjamin A. The impact of collective bargaining laws on student achievement: Evidence from a New Mexico
natural experiment. The Yale Law Journal, 120:1130–1191, 2010–2011.
[81] Lleras-Muney, Adriana. Were compulsory attendance and child labor laws e↵ective? an analysis from 1915 to 1939.
Journal of Law and Economics, XLV:401–435, October 2002.
[82] Lockard, Duane. Toward Equal Opportunity: A Study of State and Local Antidiscrimination Laws. MacMillian, New
York, 1968.
[83] Loughran, Miriam E. The historical development of child-labor legislation in the United States, June 1921.
[84] Luker, Kristin. Abortion and the Politics of Motherhood. University of California Press, Berkeley, 1985.
[85] Lumsden, Keith and Craig Petersen. The e↵ect of right-to-work laws on unionization in the United States. Journal of
Political Economy, 83(6):1237–1248, December 1975.
[86] MacCoun, Robert J. Drug War Heresies: Learning from Other Vices, Times, and Places. Cambridge University Press,
Cambridge, 2001.
[87] Maddox, Russel W. and Robert Furqay. State and Local Government, page 456. D. Van Nostrand, Princeton, 1966.
[88] Manning, Lucy and Norene Diamond. State child-labor standards: A state-by-state summary of laws a↵ecting the
employment of minors under 18 years of age. In Child Labor Series, volume 2. U.S. Department of Labor, Washington,
D.C., July 1946.
[89] Manning, Lucy and Norene Diamond. State child-labor standards: A state-by-state summary of laws a↵eccting the
employment of minors under 18 years of age. In Child Labor Series, volume 98. U.S. Department of Labor, Washington,
D.C., January 1949.
[90] Marchman, Patrick. “little NEPAs”: State equivalents to the National Environmental Policy act in Indiana, Minnesota,
and Wisconsin, 2012.
[91] Marijuana Policy Project. The twenty-three states and one federal district with e↵ective medical marijuana laws.
http://www.mpp.org/assets/pdfs/library/MMJLawsSummary.pdf, July 2014.
[92] Maryland Department of the Environment. States adopting California’s clean car standards.
http://mde.maryland.gov/programs/Air/MobileSources/CleanCars/Pages/states.aspx.
[93] Meritt, Ella Arvilla. Child labor under the N.R.A. as shown by employment certificates issued in 1934, 1935.
[94] Merritt, Ella Arvilla. Trend of child labor, 1937 to 1939. Monthly Labor Review, 50:450–467, January 1940.
[95] Merritt, Ellla Arvilla. Trend of child labor, 1939 to 1942. Monthly Labor Review, 56:450–467, 1943.
[96] Merz, Jon F., Catherine A. Jackson, and Jacob A. Klerman. A review of abortion policy: Legality, medicaid funding,
and parental involvement, 1967-1994. Women’s Rights Law Review, 17(1):1–61, 1995.
[97] Mooney, Christopher Z. Modeling regional e↵ects on state policy di↵usion. Political Research Quarterly, 54(1):103–124,
2001.
8
[98] Murray, Pauli. States’ laws on race and color, 1950.
[99] Caitlin Knowles Myers. Young women’s access to abortion and contraception, 1960-present. Middlebury College and
IZA, June 2012.
[100] NARAL Pro-Choice America. State governments.
http://www.prochoiceamerica.org/government-and-you/state-governments/.
[101] National Apartment Association. Rent control laws in the 50 states. http://www.naahq.org/sites/default/files/
naa-documents/government-affairs/protected/affordable-housing/Rent-Control-Chart.pdf.
[102] National Center for Air Pollution Control. A Digest of Air Pollution Laws: 1967 Edition. U.S. Department of Health,
Education, and Welfare, Washington, 1967.
[103] National Child Labor Committee. Child Labor Facts, Publication No. 343. National Child Labor Committee, New
York, April 1928.
[104] National Conference of State Legislatures. Emergency contraception state laws.
http://www.ncsl.org/research/health/emergency-contraception-state-laws.aspx#STLW, August 2012.
[105] National Conference of State Legislatures. Tuition benefits for immigrants.
http://www.ncsl.org/research/immigration/tuition-benefits-for-immigrants.aspx, July 2014.
[106] National Conference of State Legislatures. State policies in brief: Emergency contraceptiono.
http://www.guttmacher.org/statecenter/spibs/spib_EC.pdf, January 2015.
[107] National Conference of State Legislatures (NCSL). Equal pay for equal work – summary.
http://www.ncsl.org/research/labor-and-employment/equal-pay-laws.aspx.
[108] National Conference of State Legislatures (NCSL). States with moment of silence or school prayer legislation.
http://www.ncsl.org/research/education/moment-of-silence-and-school-prayer.aspx, 2008.
[109] National Conference of State Legislatures (NCSL). State medical marijuana laws.
http://www.ncsl.org/research/health/state-medical-marijuana-laws.aspx, January 2015.
[110] National Council of State Legislatures (NCSL). Civil unions & domestic partnership statutes.
http://www.ncsl.org/research/human-services/civil-unions-and-domestic-partnership-statutes.aspx#DC,
November 2014.
[111] National Council of State Legislatures (NCSL). State pharmaceutical assistance programs (subsidies and discounts for
seniors, disabled, uninsured, and others).
http://www.ncsl.org/research/health/state-pharmaceutical-assistance-programs.aspx, March 2014.
[112] National Education Association. A Manual on Certification Requirements. National Education Association,
Washington, 1957.
[113] National Education Association. Summary of state statutory provisions regarding scope of collective bargaining, 2004.
[114] National Education Association. Collective bargaining rights for education employees in the United States, 2005.
[115] National Gay and Lesbian Task Force. State laws prohibiting recognition of same-sex relationships.
http://www.thetaskforce.org/static_html/downloads/reports/issue_maps/samesex_relationships_5_15_13.pdf,
May 2013.
[116] National Gay and Lesbian Task Force. State nondiscrimination laws in the U.S.
http://www.thetaskforce.org/downloads/reports/issue_maps/non_discrimination_6_13_color.pdf, June 2013.
[117] National Governors Association. MCH update, 1990–2012.
[118] National Governors Association. Clean electricity: Public benefits funds.
http://www.nga.org/files/live/sites/NGA/files/pdf/1008CLEANENERGYELECTRICITYFUNDS.PDF, 2010.
[119] National Highway and Traffic Safety Administration. Determine why there are fewer young alcohol-impaired drivers:
States with the largest declines in young drinking driver fatal crsah involvements, 1982-1998.
[120] National Institute on Alcohol Abuse and Alcoholism (NIAAA). Retail sales: Keg registration.
https://alcoholpolicy.niaaa.nih.gov/Keg_Registration.html.
9
[121] National Oceanic and Atmosphere Adminstration (NOAA). Cooperation with states: ESA section 6 program.
http://www.nmfs.noaa.gov/pr/conservation/states/.
[122] Neale, Thomas H. The proposed equal rights amendment: Contemporary ratification issues.
http://www.fas.org/sgp/crs/misc/R42979.pdf, May 2013.
[123] Neumark, David. Age discrimination legislation in the United States. Contemporary Economic Policy, 21(3):297–317,
2003.
[124] Neumark, David and Wendy A. Stock. Age discrimination laws and labor market efficiency. The Journal of Political
Economy, 107(5):1081–1125, 1999.
[125] Neumark, David and Wendy A. Stock. The labor market e↵ects of sex and race discrimination laws. Economic Inquiry,
44(3):385–419, 2006.
[126] Nightingale Alliance. Legal status of assisted suicide/euthanasia laws in the United States.
http://www.nightingalealliance.org/pdf/state_grid.pdf.
[127] NORML. States that have decriminalized.
http://norml.org/aboutmarijuana/item/states-that-have-decriminalized.
[128] O’Toole, Randal. The planning tax: The case against regional growth-management planning. Policy Analysis, 606,
December 2007.
[129] Overbeck, Wayne and Genelle Belmas. Major Principles of Media Law, 2012 Edition. Wadsworth, Boston, 2012.
[130] Patients Rights Council. Assisted suicide laws in the United States.
http://www.patientsrightscouncil.org/site/assisted-suicide-state-laws/, February 2015.
[131] Barbara Perry. Hate and Bias Crime, volume 2. Praeger Publishers, 2003.
[132] Pierce, Patrick A. and Donald E. Miller. Variations in the di↵usion of state lottery adoptions: How revenue dedication
changes morality politics. Policy Studies Journal, 27(4):696–706, 1999.
[133] Je↵rey A. Ra↵el. Historical Dictionary of School Segregation and Desegregation: The American Experience.
Greenwood Press, Westport, CT, 1998.
[134] Ramseur, Jonathan L. The regional greenhouse gas initiative: Lessons learned and issues for policy makers.
http://fas.org/sgp/crs/misc/R41836.pdf, November 2014.
[135] Reporters Committee for Freedom of the Press. The reporter’s privilege. http://www.rcfp.org/reporters-privilege,
2003.
[136] Reproductive Rights. Mandatory delays and biased counseling for women seeking abortions. http:
//reproductiverights.org/en/project/mandatory-delays-and-biased-counseling-for-women-seeking-abortions,
September 2010.
[137] Scheve, Kenneth and David Stasavage. The political economy of religion and social insurance in the United States,
1910-1939. Studies in American Political Development, 20:132–159, 2006.
[138] Sirico, Louis J. Automobile lemon laws: An annotated bibliography. Loyola Consumer Law Review, 8(1):Article 15,
1995.
[139] Smith, Florence P. Labor laws for women in the states and territories: Hours, home work, prohibited or regulated
occupations; seats; minimum wage. U.S. Department of Labor, Bulletin of the Women’s Bureau, 98, 1932.
[140] Smith, Florence P. Labor laws for women: Hours, home work, prohibited or regulated occupations, seats, minimum
wage. U.S. Department of Labor, Bulletin of the Women’s Bureau, 144, 1937.
[141] Smith, Florence P. State minimum-wage laws and orders: An analysis. U.S. Department of Labor, Bulletin of the
Women’s Bureau, 167, 1939.
[142] Social Security Administration. Social Security Bulletin, 1936-2014.
[143] Social Security Administration. Quarterly Public Assistance Statistics. Various Years.
[144] State Higher Education Executive Officers Association. Figures: Public Postsecondary Enrollment, Net Tuition
Revenue and Educational Appropriations per FTE, 1988-2013.
10
[145] Stigler, George J. and Claire Friedland. What can regulators regulate? The case of electricity. Journal of Law and
Economics, 5:1–16, October 1962.
[146] Sustainable Electronics Initiative. Summary of U.S. state laws on electronic waste and disposal bans.
http://www.sustainelectronics.illinois.edu/Publications/TN10-005.pdf, June 2012.
[147] Tax Foundation. Facts and Figures on Government Finance. Tax Foundation, New York, 1941–2005.
[148] The Council of State Governments. Merit systems in the states, April 1941.
[149] The Council of State Governments. Retirement system for state employees. In Book of the States. 1950, 1952, 1956,
1958.
[150] The Council of State Governments. Occupational Licensing in the States, pages 78–80. Chicago, IL, 1952.
[151] The Council of State Governments Knowledge Center. Book of the states archive: 1935-2009.
http://knowledgecenter.csg.org/kc/content/book-states-archive-1935-2009.
[152] The Urban Institute. Welfare rules databook, 1999-2013. http://anfdata.urban.org/wrd/databook.cfm.
[153] Treadwell, Lujuana Wolfe and Nancy Wallace Page. Equal rights provisions: The experience under state constitutions.
California Law Review, 65(5):1086–1112, September 1977.
[154] Umbeck, Nelda. State legislation on school attendance and related matters. U.S. Department of Health, Education, and
Welfare Circular, 615, January 1960.
[155] United States Census Bureau. Statistical Abstracts. http://www.census.gov/prod/www/statistical_abstract.html,
1878-2012.
[156] United States Commission on Civil Rights. Freedom to the free: Century of emancipation, 1963.
[157] United States Department of Labor. Wage and hour division. http://www.dol.gov/whd/state/dollar.htm, January
2015.
[158] United States Department of Labor Employment & Training Administration. Temporary disability insurance.
http://workforcesecurity.doleta.gov/unemploy/pdf/temporary.pdf.
[159] United States General Accounting Office. Auto insurance: State regulation a↵ects cost and availability, August 1986.
[160] University of Alabama. Damages cap table. http://www.soph.uab.edu/.
[161] U.S. Bureau of Education. Compulsory school attendance. Bulletin of the U.S. Bureau of Education, 1914.
[162] U.S. Department of Health, Education, and Welfare. The state and publicly supported libraries.
[163] U.S. Department of Health, Education, and Welfare. The State and Education. U.S. Department of Health, Education
and Welfare, Washington, D.C., 1955.
[164] U.S. Department of Labor. Child Labor: Facts and Figures, Publication No. 197. Bureau of Labor Standards,
Washington, D.C., 1930.
[165] U.S. Department of Labor. Significant provisions of state unemployment compensation laws, 1937–2012.
[166] U.S. Department of Labor. State minimum-wage laws and orders: July 1, 1942-January 1, 1949. Bulletin of the
Women’s Bureau, 227, 1949.
[167] U.S. Department of Labor. State minimum-wage laws and orders: July 1, 1942-January 1, 1950. Bulletin of the
Women’s Bureau, 227 – Revised, 1950.
[168] U.S. Department of Labor. State minimum-wage laws and orders: July 1, 1942-March 1, 1953. Bulletin of the Women’s
Bureau, 247, 1953.
[169] U.S. Department of Labor. State workmen’s compensation laws: a comparison of major provisions with recommended
standards. U.S. Department of Labor Bulletin, 161 – Revised, December 1961.
[170] U.S. Department of Labor. State anti-injunction laws: A brief discussion of major provisions. U.S. Department of Labor
Bulletin, 250, September 1962.
11
[171] U.S. Department of Labor. Analysis of coverage and wage rates of state minimum wage laws and orders. Bulletin of the
Women’s Bureau, 291, August 1965.
[172] U.S. Department of Labor. State Child Labor Standards, Bulletin 158. Bureau of Labor Standards, Washington, D.C.,
1965.
[173] U.S. Office of Education. Compulsory school-attendance provisions a↵ecting employment of minors in the United
States, 1936. Monthly Labor Review, 44:356–358, 1937.
[174] Valetta, Robert G. and Richard B. Freeman. The NBER public sector collective bargaining law dataset. In Richard B.
Freeman and Casey Ichniowski, editor, When Public Employees Unionize. NBER and University of Chicago Press,
Chicago, 1988.
[175] Vernick, Jon S. and Lisa M. Hepburn. State and federal gun laws: Trends for 1970-99. In Jens Ludwig and Philip J.
Cook, editors, Evaluating Gun Policy: E↵ects on Crime and Violence, pages 345–411. The Brookings Institution,
Washington, D.C., 2003.
[176] Wagenaar, Alexander C. Alcohol policies in the United States: Highlights from the 50 states.
http://www.impacteen.org/generalarea_PDFs/Alcohol%20Policies%20in%20the%20United%20States.PDF, 2000.
[177] Wholesale Solar. Solar incentives by state. http://www.wholesalesolar.com/states.html.
[178] Yost, Nicholas C. NEPA’s progeny: State environmental policy acts. News and Analysis, 3(1), 1973.
12